Antonio Giustozzi and Mohammad Isaqzadeh Afghanistan’s Paramilitary Policing in Context The Risks of Expediency EXECUTIVE SUMMARY Despite representing the bulk of Afghanistan’s post‐2001 policing, the paramilitary dimension has received little attention among analysts. Still, if the current regime is to survive the transition to Afghan leadership in the security sector, which officially started in July 2011, and the subsequent Western disengagement, getting paramilitary policing into a functional state will be decisive. There are different types of para‐militarism, however, and what suits the Afghan predicament can only be determined by looking at the functioning of the MoI. Afghanistan had a functioning policing system before the long series of wars which started in 1978. Although that system was not particularly sophisticated, nor particularly effective, it did not look bad in comparison to the policing systems prevailing in the rest of South Asia. By the 1970s, Afghanistan’s police functioned together with the sub‐national administration as a dispute‐
settlement mechanism, as well as an early warning system, meant to detect big disturbances before they became unmanageable. The thin layer of policemen in the rural areas allowed little more than that, although in the cities civilian policing was gradually growing. By the early 1970s, it still fell short of the standard required for prosecution according the principles of the rule of law, but seemed capable of getting there in the foreseeable future. In the rural areas it remained dependent on the army for handling any serious disturbance, however, and its paramilitary dimension was mainly focused on border control. Political loyalty (to the monarchy or to President Mohammad Daud from 1973 onwards) remained a key consideration in making appointments, but took merit into consideration, particularly in Daud’s time. This allowed for a gradual increase in the effectiveness and capability of the system. The start of armed conflict in 1978 had a massive impact on policing in Afghanistan. The initial disruption caused by political purges in 1978–80 was offset by two other developments: massive investment in human resources throughout the 1980s and the much heavier politicisation caused by the introduction of party organisation within the MoI. This had the side effect of creating a system of oversight, with party members reporting through their own channels about the behaviour of their colleagues. While this smelled of totalitarianism, particularly to external observers, it kept corruption at low levels and discipline at high levels despite a very challenging situation. 2 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
The most important development of the 1980s, however, was the para‐militarisation of the police force. The police grew larger than the army, was heavily equipped and became a cornerstone of the factional balance of power within the regime. During the 1990s, however, the police lost almost all of the capabilities it had slowly accumulated over the previous 70 years. The political purges were nothing new, but no effort was made to train or select as a replacement qualified officers and the MoI became just one of many spoils being distributed. In some areas of the country, parts of the policing system stayed in place, particularly in the north under General Dostum. In the MoI itself, the purges and the exodus went so deep that when the Taleban took over Kabul in 1996 they had to re‐recruit former senior officials who had served in the 1980s to re‐establish a modicum of functionality. At the end of 2001 and in early 2002, the MoI looked quite like it had from1992 to 2006, not least because many of the people who had been appointed in those years were re‐appointed. There was however a sprinkling of old professionals in the system; their numbers went slowly up after 2002, even though only a small minority of those qualified for the job and available on the job market were brought back. Again the MoI was being used for the distribution of spoils and as a result its level of effectiveness was negligible. Pressure from donors started mounting in 2003, as presidential and parliamentary elections approached and the need for an MoI capable to secure them was increasingly felt; then the insurgency started picking up in 2005 with the police at the forefront of the counter‐insurgency effort, highlighting the need for a more effective MoI. The MoI had to grow into something more than a mere patronage system. The decision to focus the army on external defence, although not fully implementable, created space for the resurgence of para‐militarism, not just because of the insurgency but also because of the proliferation of illegal armed groups. The debate among international patterns raged for years over which model of policing would prevail – paramilitary as advocated mainly by the US Army or professional civilian policing as advocated by the Germans and most other European countries. Eventually, the view that the paramilitary option was the most realistic prevailed. At least three main reform efforts took place in 2003–5, 2006 and 2008–10 under three different ministers, but with limited results. Although at various points a AAN Thematic Report 07/2011 debate occurred on the formation of a paramilitary wing within the MoI, para‐militarisation mainly occurred as short term expediency to deal with a deteriorating security situation. By 2011, the MoI was picking individuals deemed to be particularly committed to fighting the insurgents from the same old pool of former comrades in arms and political appointees; no system of meritocratic promotion from the ranks of a new generation of senior police officers was in place or being effectively developed. Most of the efforts to improve the MoI as a system were dedicated to the command‐and‐control structure, a sound choice from the perspective of fighting an insurgency. Technologically, the improvements were massive, with communications greatly enhanced; but, the continued staffing of many positions with strongmen and factional leaders made the field units responsive to orders from the top only to a limited degree. Reporting to the centre was weak, leaving the MoI often in the dark about what was really going on in the provinces. Indiscipline was common in the ranks, sometimes in such extreme forms that it reduced the ability of field units to function to virtually nil. The focus on command and control did not derive from a coherent strategy of developing the paramilitary dimension of the MoI. The formation of a gendarmerie (ANCOP) was in progress from 2008 onwards, but on a relatively small scale and with many contradictions. The issue of what model of policing the MoI was striving for remained side‐
lined. The MoI as of 2011 seemed to be striving to re‐establish the functionality of its structure based on the model existing in the 1970s, but with on‐
going debates within the MoI and between foreign advisers about the specific roles of paramilitary and civilian policing, centralised and decentralised policing, and various types of oversight over the police. The priority was being given to state security, but with many contradictory concessions to local policing (even before the formation of the Afghan Local Police in 2010), which in theory meant to meet the security needs of the village communities. The highly centralised model was further at odds with the inability of the MoI to mobilise sufficient human resources to staff it properly; control from the top could not function effectively without multiple levels of bureaucratic supervision, which in turn required educated and skilled professionals, who simply did not exist in sufficient numbers within the structure. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context Some of the provincial strongmen who dominated much of the periphery in 2002 managed to run the provincial police systems more effectively than the MoI, as in the cases of Herat under Ismail Khan and Balkh under Mohammed Atta. As of 2011 however, there seemed to be little appetite within the MoI for seriously considering alternative models, which could range from local (‘civil society’ or local authorities) oversight, to openly letting strongmen take ownership of the police in the provinces. Centralisation remained the imperative, with neither the human resources to implement it effectively, nor a coherent plan to develop those resources. In particular, little or no effort went into improving meritocracy within the MoI, which could have potentially led to significant gains in command and control, and quickly. In sum, by the summer of 2011, the MoI still lacked a coherent strategy to bring its paramilitary capacities in line with the demands of a transition towards Afghan leadership in the security sector, aside from mere quantitative growth. Reliance on a few charismatic fighters to lead the counter‐
insurgency effort amounted to expediency more than to a forward looking plan. GLOSSARY ABP Afghan Border Police ANA Afghan National Army ANAP Afghan National Auxiliary Police ANCOP Afghan National Civil Order Police (gendarmerie) ANP Afghan National Police, incorporating ABP, AUP and others APIRS Afghan Police Incident Reporting System Khalq one of the two main wings of the PDPA, the pro‐Soviet groups in power from 1978–92 Khan big local notable, usually a landlord with a retinue of followers Kotwali the directorate of rural policing in the 1920s LOTFA Law and Order Trust Fund for Afghanistan MoI Ministry of Interior Mujahedin those who fight in the jihad; also used to indicate the veterans of the anti‐Soviet resistance NTM‐A NATO Training Mission– Afghanistan Parcham one of the two main wings of the People’s Democratic Party of Afghanistan, the pro‐Soviet group in power in 1978–92 APPF Afghan Public Protection Force, a department of the MoI AUP Afghanistan Uniformed Police CID Criminal Investigation Department CSTC‐A Combined Security Transition Command – Afghanistan IDLG Independent Directorate of Local Governance PDPA International Assistance Security Force People’s Democratic Party of Afghanistan Tashkil personnel charts UNAMA United Nations Assistance Mission to Afghanistan UNDP United Nations Development Fund ISAF Jamiat‐i Islami Afghanistan Islamic Society, an Islamist party that played a key role in the 1980s and 1990s November 2011 3
4 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
CONTENTS 1. 2. 3. 4. 5. 6. 7. Introduction How Afghanistan was policed before 2001 2.1 Before the wars 2.2 The impact of the wars (1978–92) 2.3 The impact of the civil wars (1992–2001) The police in 2002 The police after 2002 4.1 Earlier reform efforts 4.2 Towards centralised policing 4.3 Combating corruption 4.4 The rule of law 4.5 What has been achieved? Continuity and change in command and control 5.1 The command structure 5.2 Reporting to the centre 5.3 Control: Collaboration with the enemy 5.4 Control: Indiscipline in the ranks 5.5 Control: Internal affairs 5.6 Conclusion Herat: A case study of a patrimonial command and control 6.1 Police during Ismail Khan’s governorship (2001–4) 6.2 Police in post–Ismail Khan Herat Conclusion 4 6 6 9 11 12 13 13 14 16 17 17 18 19 22 23 24 26 27 28 28 31 34 1. INTRODUCTION By 2011, Western involvement with Afghanistan’s police was running into its eleventh year. Although the real take‐off in terms of investment in the police was in 2008, American investment had already picked up considerably after 2004, with US$1.1 billion spent in 2005–6. If measured against Afghan GDP (about US$15.6 billion in 2010) or against South Asian police budgets, that was a considerable amount.1 Only in 2010, however, did the limited impact of the money spent arouse serious worries among donors. During that year, transition to Afghan leadership in the security sector became the main obsession of Western diplomats and policy makers in and around Afghanistan. To some, it was the hope of a light at the end of a tunnel, ending an unwelcome engagement in a war in a distant country, away from their primary concerns. To others, it was a nightmare of uncertainty and possibly the unravelling of years of gains (whether claimed or 1
India spent about US$12 billion in 2009 for all its police forces, in a country of over 1 billion people. AAN Thematic Report 07/2011 real) in the development of the Afghan security forces. The first view predominated in Europe, while the second predominated in the United States, particularly within the armed forces. Undoubtedly, with regard to the viability of an Afghan state, within a context of decreasing Western military presence and perhaps eventually no such military presence, concerns about the state of the Afghan armed forces are justified. This paper focuses on Afghanistan’s police and the Ministry of Interior (MoI) in relation to the transition: how ready are the police and the MoI for transition and what is their future potential for assuming greater responsibilities? The commander of ISAF, General Petraeus, stated in late 2010 that transition will mean ‘we do a little bit less, the Afghans do a little bit more’.2 However, Petraeus’ views were not necessarily shared by his European partners, or even by his political masters in Washington. The speed of the process is thus likely not to be determined by military and technical considerations alone, but by political ones too. Even the Afghan government seems inclined to accelerate the transition, to regain political capital among the Afghan population and demonstrate its ability to lead the country autonomously. At the middle and lower levels of the Afghan security apparatus, however, leaders have little enthusiasm for having to fight increasingly on their own. Perhaps they know something their political masters are not aware of (or choose to ignore). Much of the public discussion on the effectiveness of Afghanistan’s police has been focused on their tactical capabilities, that is, how capable each unit is to fight the insurgents and establish control over its portion of Afghan territory. In fact, until 2010 the training and mentoring effort led by the Americans essentially focused on these tactical aspects. This paper, however, adopts a different focus: it explores whether, once the police get out of the shadow of Western armies, it will be able to fulfil its main task of keeping the existing political regime in place. Since the Afghan government is being challenged militarily, the paramilitary dimension of policing in Afghanistan will be crucial to the success of the transition: it will also be crucial to the survival of the present regime in Afghanistan, regardless of the fate of the current ‘transition process’ as such. The gradual disengagement of the Western powers from Afghanistan is probably irreversible. It deserves, 2
Phil Stewart and David Alexander, ‘Petraeus Identifies Areas for Afghan Transition’, Reuters, 19 September 2010. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context therefore, to be the focus of this paper. In turn, the viability of paramilitary policing rests on the strength of its command‐and‐control structure. A civilian police force does not need strong command and control, as long as effective forms of oversight exist. The contrary is true of paramilitary police. communicate orders to its subordinates and verify that they are obeyed. These abilities have both a technological dimension and a bureaucratic one. They also derive from the authority and the legitimacy of the leader, as well as from the leaders’ willingness to exercise them. Paramilitary policing is characterised by military discipline and military training, sometimes in combination with specialist policing training, but not necessarily. A paramilitary police is however not a small army, but a hybrid force which lacks the ability of an army to mass numbers and firepower in a specific place and time. It does not have the logistics and the command‐and‐control capabilities to do so. Instead, it can manage and coordinate, at most, hundreds of combatants in a single operation, which is still more than a civilian police force can do. Its focus is in fighting off organised challenges to state power, hence its particular relevance to the counter‐insurgency in Afghanistan. Another characteristic of a paramilitary police force is its centralised character: the military dimension requires a degree of centralisation, without which coordination would not be possible. Centralisation means that the paramilitary police is, in principle, controllable from a central command, or at least that it is designed for this to be possible. A paramilitary police force is therefore always more centralised than a civilian police force, although the actual extent of centralisation varies from context to context – in the same way that armies around the world do not all have the same degree of capacity to exercise centralised command and control. The report focuses on the current status of the chain of command and control within the MoI and on the issue of whether, realistically, it will strengthen over the next few years. It explores whether a coherent effort is being made to eliminate existing flaws, or whether the policy of the MoI is instead being driven by expediency. It makes the point that developing a paramilitary force capable of containing the insurgency cannot be isolated from the wider issue of how different policing tasks are balanced within the Afghan MoI: paramilitary policing has to be seen within the context of how the police cater for the state’s need for self‐protection, the demand of communities for basic security, the desire of the business community and other groups for predictability in policing, and a range of other tasks. The central question is whether the Afghan state as it is can survive the transition without developing a relatively balanced package of policing. Centralisation is the concentration of the power to make decisions at the top of an organisation and the capacity to have them implemented at the bottom. Various degrees of centralisation can be achieved in a number of ways, not all of them easily compatible with para‐militarism. Ideally leadership can have its orders implemented at the bottom by being able to sanction those not complying. If applying sanctions is not possible for bureaucratic or political reasons, a surrogate option is to appoint individuals judged to be personally loyal to the leadership in positions of power. The weaknesses of this surrogate are at least three: enough qualified and loyal candidates may not be available; judging the loyalty of individuals is not always easy; and finally, even loyal individuals might be reluctant to implement orders which go against their interest, if there is no risk of sanction. One particular aspect of centralisation is command and control, that is, the ability of leadership to For this reason, this paper does not go straight to the issue of command and control, but starts by showing how the police force developed from its origins to 2001. This is discussed in Section 2. Section 3 provides a snapshot of the police in 2002, the point of departure of successive efforts to rebuild it, while Section 4 assesses developments within the police force after 2002. All this serves the purpose of illustrating the role of the paramilitary component within Afghanistan’s policing; how it grew and changed over time in response to varying circumstances. One additional, external dimension of the context in which paramilitary policing operates is the role assigned to the military: when it was greater (1930–70s), the police did not need to take over much of a paramilitary role; but, as the army was asked in 2002 to confine itself to an external role, the need for paramilitary policing became more and more compelling. This aspect is also discussed in the different sections of the paper. The bulk of the report is dedicated to the various dimensions of command and control and discusses various aspects of it. Section 5 focuses on the command structure and responsiveness of the units on the ground, reporting back to headquarters (HQ), the extent of indiscipline within the ranks, and the effectiveness of the internal affairs function and logistics. The paper November 2011 5
6 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
concludes by zooming in to a specific in‐depth study of Herat’s police, which illustrates how the exercise of command and control has played out in a particular province. Although Herat is not necessarily representative of other Afghan provinces, its case illustrates the shape command and control can take, away from the capital. Because of the focus on paramilitary policing, this paper does not discuss various other aspects of policing in Afghanistan in great detail. Among these are the reform efforts which succeeded each other after 2001, the various experiments with auxiliary and local policing, criminal investigation, counter‐narcotics, etc. In some cases these aspects have been covered in other publications, although some remain neglected to this day.3 2. HOW AFGHANISTAN WAS POLICED BEFORE 2001 This section will review the development of Afghan policing before the current political regime was established in 2001. Several periods can easily be identified, each with specific characteristics. The monarchy and Daud’s republic were characterised by continuity, even if an acceleration of investment in the police could be detected after 1973. The start of the period of conflict in 1978 represented a major disruption of this continuity; the leftist regimes of 1978–92 consciously tried to reshape policing according to their ideological views, even if from the bureaucratic point of view little changed. The Mujahidin republic (1992–96) did not dedicate much attention to policing and mostly used the police to distribute spoils among political stakeholders. The Taleban regime (1996–2001) implemented a model of policing, although a very basic one, resembling in some ways the one utilised in the early days under Abdur Rahman (1880–1901). 2.1 Before the wars The Ministry of Interior (MoI) as such was established in Afghanistan under King Amanullah in 1299 (1920–21). During this period, the MoI mostly dealt with sub‐national administration and did not have a policing role.4 Policing pertained instead to the Public Security Ministry, which was divided into two main directorates, urban police and rural police (kotwali), with the former concentrating on investigative functions and urban policing functions such as traffic, and the latter looking more like a paramilitary organisation. This ministry also incorporated political policing and intelligence functions. Training took place in the Police and Kotwali School and in the Kotwali Cadet School, located in Kabul.5 Under Nadir Shah (1929–33) the Public Security Ministry was incorporated into the Ministry of Interior under the name Security Department.6 During the 1930s, German and Turkish advisers and instructors started arriving, with the former leaving in 1941 but returning in the 1950s, while the Turks maintained a more permanent presence, leaving behind the influence of their own strong centralist traditions.7 The ability to exercise direct command and control of the MoI structures in the provinces was modest due to technical limitations; only in 1350 (1971–2) the MoI acquired 21 fixed wireless communication sets and a mobile one from West Germany.8 Horses, cars and other vehicles were available in limited numbers and mostly in the rapid‐
intervention battalions. The HQ was better equipped, but each province would not have had more than one or two vehicles, mostly used by the chief of police.9 3
See Eva Gross, Security Sector Reform in Afghanistan: the EU’s Contribution (Paris: EU Institute for Security Studies, 2009); Royal United Services Institute for Defence and Security Studies and Foreign Policy Research Institute, Reforming the Afghan National Police (London: Royal United Services Institute for Defence and Security Studies, 2009); Tim Foxley, ‘Where Are We with the Afghan Police Force?’ European Security Review 43 (2009); International Crisis Group, Policing in Afghanistan: Still Searching for a Strategy, Asia Briefing No 8518 (Bruxelles, 2008). Criminal investigation is not addressed in detail by any of these studies, which tend to focus on the reform efforts. On local and auxiliary policing see ‘Village Defence: Understanding the Afghan Local Police (ALP)’ (Civil‐Military Fusion Centre, July 2011) and Mathieu Lefèvre, Local Defence in Afghanistan: A Review of Government‐Backed Initiatives (Berlin: AAN, 2010). AAN Thematic Report 07/2011 The extension of the presence of the police in the rural areas happened as more and more 4
Shah Mahmud Hasin, Wezarat‐e dakhela dar gustarda‐
ye tarikh (Kabul: State Printing House), 35 and further. 5
Hasin, Wezarat‐e dakhela, 40–1 (see FN 5); Vartan Gregorian, The Emergence of Modern Afghanistan (Stanford: Stanford University Press, 1969), 299. 6
Hasin, Wezarat‐e dakhela, 41–2 (see FN 5); interview with Sayed Mohammad Gulabzoi, Kabul, October 2007. 7
Hasin, Wezarat‐e dakhela, 54–5 (see FN 5); Scott Chilton et al, Evaluation of the Appropriate Size of the Afghan National Police Force (Tashkil) (Kabul: IBF, 2009), 25. 8
Hasin, Wezarat‐e dakhela , 7, 5 (see FN 5). 9
Interview with Col. Nekzad, Afghan police, October 2007. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context administrative divisions (provinces, districts and sub‐district), each having a police garrison, were established and as the staffing of individual MoI units was strengthened.10 In every province the kotwali commander was placed in charge of security matters and was hence given the title of security commander. For a while the kotwali relied on village militiamen, kotwali deh bashis, civilians wearing a particular dress, whose salary was half paid by the kotwali and half by the villagers. Soon, however they were replaced by state security personnel.11 Initially totally reliant on the army for support in the event of serious trouble, the MoI gradually developed its autonomous capacity to intervene. In 1314 (1935–6), for the first time a central reserve police consisting of volunteers was established and organised in three battalions.12 More battalions were established in the following years and eventually in 1317 (1938–9) a gendarmerie was established, incorporating the mobile reserve and also the border police.13 By the 1960s, the gendarmerie was a mobile force, mostly on horses, with a few vehicles too.14 Even after acquiring some capability for large‐scale operations, however, the MoI remained dependent on the back‐up of the army for any serious confrontation. Gendarmerie aside, the organisation of the rest of the police was essentially paramilitary too: military ranks, top officers taken from the army, and strict military discipline. All in all, the degree of control over territory and the population by the central government was still modest by the 1970s. Official statistics for 1349 (1970–1) show only a handful of murder incidents throughout the country, as well as 735 incidents of banditry and 71 of smuggling.15 Although the MoI used such statistics to illustrate its success in containing crime, they actually demonstrate how little reach the MoI had within the population. It is not credible that a country of the size and population of Afghanistan experienced such negligible levels of criminal activity; simply put, very few crimes were being reported. The tribal areas of the Pashtun belt were left to their own devices as far as crimes and disputes were 10
Hasin, Wezarat‐e dakhela , 51–3 (see FN 5). Hasin, Wezarat‐e dakhela, 41–2 (see FN 5). 12
Hasin, Wezarat‐e dakhela , 51 (see FN 5). 13
Hasin, Wezarat‐e dakhela, 54–5 (see FN 5). 14
Harvey H. Smith et al, Area Handbook for Afghanistan (Washington: Government Printing Office, 1969), 367. 15
According to Hasin there had been ‘6–8 murders’, Wezarat‐e dakhela, 76 (see FN 5). 11
concerned. The police force was however less inclined to allow the communities to handle their own justice affairs in the north; although the police cooperated with the elders, the MoI did not allow them to administer justice as in the Pashtun areas and maintained a more interventionist attitude. As it will be shown below, after 2001 the same system of letting communities administer their own justice was adopted in much of the north too, as political relations between the centre and the ethnic minorities changed.16 At the same time, the MoI was able to enforce compliance when needed and with minimal effort. Although nostalgia about the pre‐war past might paint an excessively rosy picture, everybody seems to agree that the appearance of a lone policeman armed with a stick was enough to prompt villagers into cooperation; suspects would be handed over without resistance. Even khans could be arrested peacefully, including in places like Kandahar where even in the 1970s they had retinues of armed men. The khans never challenged the police and relied instead on negotiations and political patronage to get out of prison.17 In the rural areas, this compliance was not so much the result of respect or fear of the police, but of the knowledge that resistance would lead to the intervention of the army. Sub‐district (‘alaqadar) police stations never had more than fifteen policemen and often as few as three, with a few bolt‐action rifles and pistols; they were not able to handle serious opposition. The district police stations had a maximum complement of 36. In Afghanistan’s own version of the social contract, the state demanded little in terms of control over dispute resolution, petty criminality, most forms of social disorder and even serious crime in parts of the country, in exchange for society’s tolerance of the political regime. The police were a conscript force until 1992 and were recruited through the same system utilised by the army. The recruits, who made up the bulk of the police (the patrolmen), were those the army declined to take. Recruits whose families were well connected politically often chose to serve in the police because it was less demanding than serving in the army. Under Daud, high school graduates started being sent to the police; this improved the 16
Interview with Col. Nekzad, Afghan police, October 2007. 17
Interview with Mohammad Aman Kazimi, former district governor, London, June 2008. The episode of the lone policeman with a stick arresting a culprit in a remote village was recounted in tens of interviews with both police and non‐police. November 2011 7
8 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
quality of human resources available.18 However, salaries were low: in the early 1970s, patrolmen were paid 80 US cents a month and were therefore not very motivated. The literacy rate was barely 2 per cent and even in the last years of the monarchy (early 1970s) record keeping was ‘virtually non‐existent’, according to US government sources. In the provinces ‘lack of authority, know how, initiative, and equipment reduce[d] the police virtually to a token presence’.19 But by the 1970s, at least, a system had been established which featured the basics of functional command and control. The police force was well disciplined and as a result was seen as taking its job seriously. Policemen seemed ‘to carry out their duties conscientiously in accordance with the orders of their superiors and seemed to treat high and low alike for similar infractions if they feel they have the support of their superiors’.20 Discipline among the conscripted patrolmen was good because the MoI had the authority to extend their period of service as a punishment. Even today (2011), many police officers recruited in the 1970s or earlier believe that in a system where so many patrolmen were illiterate, military discipline was necessary and good. In the 1970s, military discipline did not just exist on paper (as after 2001), but was effectively implemented by a professional officer corps.21 With its limited means, the MoI managed to maintain a degree of functionality in the system, meaning that the police force was able to achieve its tasks as determined by the political leadership of the country. By Daud’s time, patrols were enforced through a system of checkposts and logs, which certified that patrolmen had carried out their tours as required. Still, most villages saw traces of the police no more than once every 18
Personal communication with Shah Mahmood Miakhel, former deputy Minister of Interior, July 2011. 19
‘World Drug Traffic and Its Impact on US Security: Hearings before the Subcommittee to Investigate the Administration of the Internal Security Act and Other Security Laws of the Committee on the judiciary, United States Senate, Ninety‐second Congress, Second Session, Part 4, The Global Context: Report of Gen. Walt’ (Washington: US Government Printing Office, 1972), 152. 20
H. Smith et al, Area Handbook, 366 (see FN 15). 21
Interview with Col. Nekzad, Afghan police, head of crime and investigation unit of MoI, Kabul, May 2006; interviews with police officers in Kabul, October‐
November 2010; telephone interview from Vancouver with Lloyd Baron, who was in Afghanistan in the early 1970s for studies, September 2007. AAN Thematic Report 07/2011 several months and usually much less. The police would mostly visit villages when a crime or a clash was reported.22 Under President Daud Khan (1973–78) the system of internal reporting improved and at least some paperwork started being done. Senior officers were effectively controlling their subordinates and every 3–6 months a delegation would visit the provinces, to check what was going on.23 The cooperation between the governors and the police was very close, with the former being very clearly in charge of the latter.24 However, the system was fragile due the bluntness of the army as a repressive tool and the weakness of the police as an institution more specialised in internal order; in principle the thinly spread police could be more discriminate in the repression of open opposition to the government, because it had more local knowledge than the army, but it was poorly staffed (particularly in the provinces) and too thinly spread to be able to handle more than low‐scale banditry. The principle of collective responsibility (whole communities punished for the wrongdoing of individual members, often through the destruction of the village by the army or the deportation of elders) worked efficiently to keep the population under control when dissidence and crime were community‐based.25 However, it did not work as well against organised dissent that had an ideological dimension and some social base. From 1976 onwards, politicised gangs linked to the Maoists appeared in the areas north of Kabul and tested the capacity of the police, even if only tens of individuals were involved.26 In sum, by the 1970s the MoI was gradually developing some features of civilian policing in the cities (although almost entirely still under military discipline), while in the villages a police force under military discipline mainly maintained support for the civilian administration in dispute settlement. In addition, the police force played a role of early warning in the management of political dissent and major disturbances; it could 22
Interview with Kazimi (see FN 18). Interview with Col. Nikzad, MoI, October 2007. 24
Interview with Col. Nikzad, May 2006; Interview with Lloyd Baron (see FN 22); interview with Mohammad Aman Kazimi, former district governor, London June 2008. 25
On Abdur Rahman’s use of the system see Hasan Kakar, Government and Society in Afghanistan (University of Texas press, 1979). 26
Nabi Azimi, Urdu wa syasat (Peshawar: Danish, 1998), 120. 23
Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context do little else than alert the government if the situation was getting out of control. The army still played an essential role as a backup. In terms of its appointments policy, the system under President Daud could be described as a ‘moderate meritocracy’,27 in that it tried to combine some meritocratic promotion with political favour, typically by trying to appoint people who were qualified for the job while also politically loyal. This meant that, while the best‐qualified candidates might not have been appointed, those taking the job in their place were not altogether incompetent. 2.2 The impact of the wars (1978–92) The long period of war (1978–2001) completely reshaped the policing system as it had developed by the 1970s and eventually destroyed it. A major restructuring of the police took place in the late 1970s and early 1980s, with the change of regime in Kabul and the coming to power of Khalqi leaders Taraki and Amin (1978–79) and then of Parchami leader Karmal (1979–86). The most important feature of this restructuring (apart from further sophistication, which paralleled the expanding size of the MoI in 1980–92), was the introduction of a specific role for the ruling party within the Ministry of the Interior. The politicisation of the police had already started during Daud’s republic (replacing nepotism based on kinship, which had the overriding role under the monarchy), when he promoted close political associates to the key positions,28 but it escalated to unprecedented heights in 1978–9. When the new regime took charge, the police force was seen as predominantly pro‐Daud and pro‐king, with just three or four Parchamis and a couple of Khalqis among the leading officers. The synchronisation of the police with the new communist regime started with the introduction of 160 party loyalists, mostly by transferring people from the army to the top positions and recruiting party members from 27
We define here ‘moderate meritocracy’ as a system where appointments and promotions are done by taking merit and capability into account, but in combination with other considerations, such as political loyalty. Typically in moderate meritocracies the promotion or appointment system is open to political interference, but the latter is constrained (for example by oversight of some kind) or self‐constrained (for example by the awareness of the political authority that too much is at stake for incompetent people to be staffing key positions). 28
Interview with Col. Nekzad, May 2006, Kabul. outside the security establishment.29 The purges and the gradual slide towards civil war led to defections from the police and declining recruitment; by late 1979 only 5–6,000 police were left, according to one estimate, compared to the 30,000 or so who were there at the peak of the force’s strength.30 The arrival of the Soviet Army at the end of 1979 and the change of the guard between Amin and Karmal was followed by a mixture of change and continuity. The MoI had to be rebuilt almost from scratch. The February 1980 riots in Kabul, started by university students opposed to the presence of the Soviet Army, demonstrated how ineffective the police had become: the Afghan army had to intervene heavy‐handedly with Soviet support and at least 90 civilian lives were lost.31 The Karmal government tried to replenish the ranks and increased the quota of recruits allocated to the MoI at the expense of the army. The MoI was also the first among the armed service to offer incentives for people to volunteer, and by 1984 about a third of the force was composed of volunteers.32 The police force was better supplied than the army and the general conditions of service were also better, not least because Interior Minister Gulabzoi managed the logistics directly and used his political weight to force other ministries to deliver.33 For the first time, the responsibilities of the MoI were reduced, with the transfer of Border Affairs to the Ministry of Defence and of sub‐national government to an independent department under the presidency, two moves probably dictated by President Karmal’s desire to weaken his rival Gulabzoi.34 West German advisers were replaced by Soviet and Warsaw Pact ones during 1979. More specialist faculties were established at the Police Academy in 1981, including criminal investigation and border policing. Students were sent abroad for further study in numbers greater than ever, particularly to the Soviet Union, but also Czechoslovakia, Bulgaria, Hungary and the German 29
Interview with former police officers Ghulam Mohammad Kamal and Khawal Shinwari, London, January 2008. 30
Interview with Kamal and Shinwari (see FN 30). 31
Azimi, Urdu (see FN 27), 245. Pro‐opposition sources placed the number of casualties at the time of the events much higher, in the hundreds. 32
A. Giustozzi, War, Politics and Society in Afghanistan, 1978–1992 (London: Hurst, 2000), 95–7. 33
Interview with Kamal and Shinwari (see FN 30). 34
Interview with Gulabzoi (see FN 7). November 2011 9
10 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
Democratic Republic.35 Inevitably the influence exercised by the Soviets further strengthened centralised decision making, entrenching the role of the ruling party in the process. The main aspect of continuity was in the politicisation of the police, although it was systematised and rationalised. This process of systematisation and rationalisation had several facets. Under Gulabzoi, the promotion of party stalwarts was subordinated to the recruitment of those possessing relevant qualifications.36 The presence of the party within the MoI did not just serve to reward followers as had been the case in 1978–79, but had two other aims. The first was to guarantee the loyalty of the police to the regime. The second was to exercise oversight to the bottom level, with a separate chain of reporting to the party leadership. Politicisation, however, also turned the police into the tool of a particular portion of society (the ruling PDPA and its sympathisers), making it more difficult for the police to engage with the rest of the Afghan population, particularly the large sections that were opposing the PDPA regime. This situation continued throughout the 1978–92 period, even though from 1980 onwards the police seem to have been perceived as less ideological than in 1978–79. The competition between the different wings of the party was another aspect of the politicisation of the MoI: Amin’s ‘Red Khalqis’ (as his supporters were known) were purged and some of the military transferred to the police were transferred back.37 The MoI turned into a stronghold of the anti‐Amin wing of the Khalqi faction under Gulabzoi. Parcham, the other main wing of the PDPA, maintained a presence in the MoI, but was on the whole marginal. The leadership of the party and the country was Parchami, thus casting the MoI as something like an internal opposition stronghold within the PDPA. 35
Shah Mahmood Hassein, Wezarat‐e dakhela (see FN 5), 116 and further; interview with Sayyed Mohammad Quddusi, police academy Kabul, October 2007; interview with Nekzad, 28 May 2006; interview with Kamal and Shinwari (see FN 30). 36
Interview with MoI official, Kabul, April 2010. While in 1980 just 5% of the Sarandoi (old name for the police) were members of the PDPA, by 1982 the number had gone up to 12% plus another 16% in the party youth; by the mid‐1980s about 35% of the MoI staff carried the membership card of the PDPA and every unit of the MoI had party organisations. By 1988, 46% of the Sarandoi had PDPA membership cards and another 20% were in the party youth. Giustozzi, War (see FN 33), table 23. 37
Interview with Kamal and Shinwari (see FN 30). AAN Thematic Report 07/2011 His former officers describe Gulabzoi as in full control of the MoI; nobody was able to challenge him and those who tried were removed or posted to ‘bad’ regions such as Panjshir or Kandahar. Always well informed and organised, he was ready to defend his men against the government and even the Soviets, in turn winning over the loyalty of most of them.38 He left little space to Soviet advisers, with whom in any case he maintained good relations. The advisers were mostly concentrated at the top levels of the MoI, with few in the provinces and none in the districts. The officers who served at that time recall a smooth relationship with the advisers, who were not perceived as invasive. They seem to have had little impact on paramilitary policing in particular, probably as the Soviet Ministry of Internal Affairs had a very limited experience in counter‐
insurgency.39 Charismatic leadership and leadership by example helped to assert Gulabzoi’s authority and establish his legitimacy within the MoI, resulting in effective command and control over his force in the 1980s, despite the difficult communications. He and his aides would share the same food as the rank and file, a display of egalitarianism which seems to have been appreciated.40 In part at least because of the radical policies and the violence of the Khalq regime in 1978–79, the Karmal (1979–86) and Najibullah (1986–92) regimes faced very unfavourable odds in their efforts to stabilise Afghanistan, in particular from 1980–85. The countryside was armed and under the influence of a variety of opposition armed groups; the constant inflow of weapons eventually allowed the armed opposition to incorporate hundreds of thousands in its ranks. The police had little contact with the people outside the cities; they were even instructed not to accept food because of some cases of poisoning.41 To bring some MoI influence to the villages, it became common practice to recruit villagers hostile to conscription or even previously aligned with the opposition by promising to let them serve in their village, although under the command of officers dispatched from the centre. The presence of professional officers leading these units set this police force apart from the later experiments in ‘local policing’ (the Auxiliary Police in 2006, the APPF in 2009 and the Local Police in 2010), where 38
Interview with Kamal and Shinwari (see FN 30). Interviews with former officers of the police, 2008–10. 40
Interview with MoI official who served under Gulabzoi, April 2010. 41
Interview with Kamal and Shinwari (see FN 30). 39
Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context no professional officers were on the ground. In this way, the MoI expanded its presence to a few thousands of villages (out of 44,000), but more than 80 per cent remained outside its control throughout the 1980s and early 1990s.42 Gulabzoi was replaced in 1988 because of his suspected disloyalty towards President Najibullah, who in the wake of the Soviet withdrawal feared a Khalqi move to take power. His successors, Watanyar and Pakteen, stayed close to Najibullah and changed little in the modus operandi of the MoI, but lacked Gulabzoi’s charisma. The police suffered under the purges of Khalqis which characterised 1990–2, as Najibullah tried to pre‐
empt a possible Khalqi coup. At the same time, the area under government control shrank further, while the pro‐government militias were given free rein, even in the cities, contributing to a deterioration of law and order.43 In sum, from the perspective of policing, in 1980–
92 period para‐militarisation became paramount – not because of the army’s reduced role, but because of the army’s inability to cope on its own, and due to rivalries internal to the regime. In some regards, the pre‐1978 attitude of ‘moderate meritocracy’ returned: merit was considered in making appointments, but so was political loyalty. Once appointments were made, strict adherence to a code of conduct was expected and enforced. The type of political loyalty and the way it was supposed to be expressed differed compared to Daud’s republic and the monarchy, the party organisation now being the cornerstone of it; however inter‐party personal and factional loyalties mattered more than simple party membership. Efforts to improve the professionalism of the police were resumed and received larger investment than ever before, but with the priority always being para‐militarisation (territorial and population control). 2.3 The impact of the civil wars (1992–2001) Inevitably, with the collapse of Najib’s regime in April 1992 and the ensuing international neglect of Afghanistan, the MoI’s huge structure developed under Karmal and Najib could not be sustained. Although the structure per se was not immediately dismantled, staffing levels collapsed. All the MoI’s activities were paralysed during this period; the 42
43
Giustozzi, War, 342 and further (see FN 33). Giustozzi, War (see FN 33). new mujahidin appointees could not even manage the release of passports.44 As a result, the MoI lost most of its professional staff. Few stayed on. The police force was used as a source of patronage and nepotism, with no consideration for merit: the number of generals proliferated in the hundreds, despite reductions in the ranks. Occasionally in the 1990s, policemen wearing uniforms could be spotted, but largely discipline was not enforced and recruitment of party or faction loyalists became the rule. Formally, the bureaucratic structure and the regulations remained, but they were rarely enforced. In areas outside Kabul’s control, either no police force existed, or local arrangements were made where local councils maintained some police force.45 Where the police continued to operate, there is no particular evidence of an increase in corruption, although sources for the period are limited. An NGO source with experience of the north commented that in Mazar‐e Sharif under Dostum the police appeared content with taking some apples from the traders and complaints of serious misbehaviour were not heard; at the same time however the police did nothing to prevent abuses by the warlords, which were frequent.46 During the Taleban regime (1996–2001) the police force was little more than a branch of the army, staffed by detachments of fighters who enforced law and order while resting from the battlefront. Whereas under President Rabbani some efforts to resume police training had been ongoing, under the Taleban training was almost completely abandoned. The police as an institution collapsed as the Taleban dismissed the bulk of what was left and placed their own fighters in police stations.47 The police stations, manned by young fighters without uniform and mainly from Pashtun regions, operated almost autonomously. The Taleban fighters occupying the police stations would not open a file or record of the petitions or the cases. They would not refer the criminal cases to the Criminal Investigation Unit of the Provincial Police Directorate or report to it. While the Islamic 44
Personal communication with former MoI officers, 2008–10. 45
Interview with high‐ranking MoI official, Kabul, May 2006. 46
Interview with Peter Schwittek, NGO manager, Kabul, October 2007. Gen Dostum was one of the key participants in the 1990s civil war and the official leader of Junbesh‐e Melli from its inception until recently. He recently resumed his job as Chief of Staff of the Commander of the Armed Forces. 47
Interviews with police officers who worked under Taleban rule, Herat, July 2009. November 2011 11
12 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
Republic, in principle, still paid respect to bureaucratic rules and procedures, the Taleban’s Islamic Emirate abandoned them completely. Most remaining professional police officers, who had received official training under Najib’s rule and who were not dismissed by the Taleban, were working at the Passport Department and ID Issuance Department. Some senior police officers continued to serve in the MoI headquarters, or were re‐hired by the Taleban to make up for the shortage of skills.48 The few police officials allowed to wear uniforms were traffic police officers were stationed on the streets to manage the flow of traffic.49 The Taleban do not appear to have used the police for distributing patronage; they simply lacked the resources to maintain a professional police force and reduced it to little more than a subsidiary of the army, as it had been before King Amanullah. 3. THE POLICE IN 2002 As Afghanistan emerged from the Taleban regime and entered a new phase of its history, little was left of any state institution. The few officials who had managed to serve all the regimes from the 1970s onwards stayed on; others who had served under Rabbani and then quit came back and re‐
occupied their posts. In the MoI, the process largely took place under the party that had received the ministry among the spoils of war, Jamiat‐i Islami. The top positions and many of the intermediate and lower positions were filled with Jamiati loyalists. Almost no equipment was left, not even chairs and desks. Neither uniforms nor vehicles were available. In these conditions, a major re‐building effort was necessary. A debate started over which model the new MoI should follow. Most Afghan players either had no particular model in mind, or were inclined to stick to those of the past: typically the MoI of the 1970s or of the 1980s, when Soviet patronage had allowed a massive expansion of the police. The donors and the intervening powers (which in part coincided) were inclined to bring Afghanistan more in line with policing as practiced in the Western hemisphere, moving towards a civilian model (as opposed to the paramilitary model prevailing in Afghanistan).50 48
Personal communication with former Deputy Interior Minister Abdul Hadi Khalid, Kabul, April 2011. 49
Interviews with police officers who worked at the provincial police headquarters, Herat, July 2009. 50
Interviews with police trainers, advisors and mentors, Kabul, 2003–2010. AAN Thematic Report 07/2011 Recruitment was voluntary and fully decentralised; any relationship with the army was severed (until the Taleban there had been a unified recruitment structure) and self‐appointed chiefs of police (usually strongmen linked to the anti‐Taleban factions) recruited officers and patrolmen as they pleased, with no oversight. The anti‐Taleban factions were however wary of supplying weapons to the police staffed by their own men (preferring instead to hoard them) and for years after the fall of the Taleban shortages of weapons was common in the police stations. A few professional police officers were brought back to handle whatever little administration was there, as well as logistics and communications; some also made it to the CID (Criminal Investigation Department). The weapons were usually worn‐out Kalashnikovs or cheap copies made outside Russia. Salaries were very low and rarely paid on time; allowances, bonuses and even part of the salaries were often embezzled in Kabul or locally by corrupt officers. The more‐
specialised components of the MoI were particularly neglected: the intelligence department’s budget was so small that it could not deploy an officer in every district.51 As a result, in 2002 and following, discipline within the police force was poor. There were exceptions, as the Herat case study below shows, where even with very limited human resources some security was delivered, depending on who was in charge at the provincial level and his ability to enforce his own disciplinary regime. The majority of the provincial strongmen turned into chiefs of police had neither the ability nor the will to set up a functioning police force. The MoI had little ability to intervene. Internal Affairs hardly operated, while investigative capability was minimal everywhere, non‐existent in some provinces. Ghost police (policemen listed in the paybooks, but fact absent from the ranks) were a widespread problem, eating away much of the meagre resources allocated to the MoI by the Ministry of Finance. Whole sections of the population were excluded from any potential benefit deriving from policing because they could hardly approach the police directly. This was in particular the case for women, who had to rely on their male relatives to contact the police, but also of most of the rural areas; without vehicles the police could not reach to them. On the other hand, given the level of police corruption, its indiscipline and incompetence, and the links between chiefs of 51
Interview with chief of police of Baghlan, Pul‐i Khumri, December 2003. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context police and strongmen or factions, not many Afghans were keen to see the police at all.52 The command‐and‐control structure of the police was very weak at the national level and relied on the top ministerial staff contacting the provincial chiefs of police and asking them to take a particular action. Even at the national level, personal relations – usually along factional lines – determined the responsiveness of the system to orders coming from the top. At the provincial level, command and control depended on the personal status of the chief of police; in areas where he had full control because he had staffed the police with his own men, he could issue orders and expect obedience in the districts (that is assuming he was able to communicate, given that in 2002 the police was not equipped with UHF radios and phones did not work). But if the police force at the provincial level had been divided among factions and strongmen as part of the division of the spoils following the collapse of the Taleban, then even the provincial chief of police would not be able to exercise much control and friction often occurred.53 At least in 2002 Afghanistan seemed to be at peace, apart from a few pockets of residual activity by Taleban and its allies and occasional factional clashes in the north. Casualties were minimal as the police hardly engaged in the ‘counter‐
terrorism’ effort. Fighting crime was not much of a concern either, not because crime did not occur, but because the police let the communities deal with it or simply tolerated the criminal activities of the strongmen and the militias to which it was connected; in fact the police force itself was a major source of criminal activity.54 Arguably, therefore, in 2002 Afghanistan’s police force did not aspire to follow any model and was largely a resource for the distribution of political patronage. In certain regards it was definitely a paramilitary force, with military ranks and a focus on fighting political dissidence, even if in other regards (weak discipline and command and control) it resembled a semi‐regular militia. Although not an exceptional situation to be in after many years of conflict, inevitably the police would at some point be called upon to perform some higher duty than simply rewarding the protagonists of the anti‐Taleban mobilisation of 2001 with 52
Interviews with UN officials, senior police officers, foreign diplomats and police advisors, 2003–10. 53
Ibid. 54
Interviews with police officers in Kunduz, Takhar, Balkh, Faryab, Herat and Kandahar, 2003–5. prestigious appointments. Indeed pressure to reform started building quite soon, as donors were keen to see some returns for the money they were pouring in and as the security environment deteriorated. 4. THE POLICE AFTER 2002 The dire state of the MoI and the police in 2002 prompted several reform efforts, not so central to the topic of this report, but relevant to determine the context within which para‐militarism has been evolving. They are discussed in Section 4.1. Section 4.2 is dedicated to efforts to re‐centralise the MoI, which despite not being particularly successful had a degree of impact. The spread of corruption is the most‐debated aspect of policing in Afghanistan and is discussed in Section 4.3; corruption has an obvious impact on the functionality of every institution and therefore affected paramilitary policing too. The issue of re‐establishing the rule of law is discussed in Section 4.4. This is where civilian policing would be expected to be clearly superior to paramilitary policing. The section shows how the slow development of paramilitary policing was not because the main effort was going towards establishing the rule of law – which was instead even more neglected. Although few of the reforms launched in this period could be described as even partial successes, some achievements were made. The balance of successes and failures is discussed in the last section, 4.5. 4.1 Earlier reform efforts An important decision taken by the government in 2002 was to specifically task the army with external security, leaving internal security to the MoI.55 In practice, in subsequent years this could not be fully honoured, as the expanding insurgency forced the army to intervene in internal security, but the decision was crucial in creating the need for a large paramilitary force. At no point during all these debates on reform and attempts to 55
The initial version of the Afghan National Military strategy stressed the defence of the borders and national integrity: see OMC‐A, ‘First Afghan Security Strategy Signed, Defense and Security Highlights’ (Afghanistan, 6 November 2004). The 2010 version of the strategy was somewhat more ambiguous: according to the Afghan National Military Strategy of 15 February 2010, the Afghan National Army is ‘responsible for protecting Afghanistan’s territorial integrity, upholding and protecting the Constitution, defending national and Islamic values, and establishing a favourable environment for public welfare and progress’. November 2011 13
14 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
implement it was the predominant paramilitary character of Afghanistan’s police in doubt; even Atmar’s plan (see below) would probably have left the civilian police as a comparatively small component of the force. Several efforts to reform the MoI occurred after 2001, driven by part of the central government’s desire to turn the MoI into a tool for rebuilding a centralised Afghan state. The first round of reforms, during the tenure of Minister Jalali in 2003–5, was mainly about establishing a more functional structure, strengthening control over the periphery and improving the appointments process. The results were mostly limited to structural change, with a rationalisation of the departmental organisation. Political opposition to stronger control from the centre and a more limited role of political patronage in recruitment prevented greater gains. While reform efforts can have important positive consequences in the long run, they are destabilising in the short run. In the case of Afghanistan, reform efforts had the negative side effect of creating instability, by making MoI officials fearful of losing their jobs; corruption in all its forms might well have increased as a result, helped because efforts to increase oversight, accountability and internal controls largely failed (see also Section 4.3 below).58 Moreover, in some provinces where a policing system was delivering a degree of security if not the rule of law, the reform effort undid it by eliminating the local charismatic leadership, without establishing a functioning alternative system (at least not yet). This was the case in Herat, as discussed in Section 6. 4.2 Towards centralised policing A third reform effort took place in 2008–10 under Minister Atmar, again with strong international input. This effort met with some success, as during Jalali’s tenure, in structural rationalisation, but little in the fight against corruption and political interference in appointments. Atmar was the first to develop a plan to overhaul Afghanistan’s police, proposing the formation of three distinct forces: a gendarmerie, a civilian police force and a local police force, to protect the communities. However, he fell short in implementing the plan as he had little support within and outside of the MoI.57 Although centralisation is not synonymous with para‐militarisation, the latter requires the former. A collection of effective and charismatic leaders does not make an army (or a paramilitary force); what is required, as explained in the introduction, is the capability to act in a coherent and integrated way. The slow process of centralisation of policing after 2001 impeded successful para‐militarisation, both before and after it became the main effort within the MoI. On the other hand, the issue is not merely about centralisation, as centralised policing can take different shapes and have different purposes. Centralised policing may help to consolidate the state, if the latter has the management capacity to handle it, but is less likely to serve the interests of local communities, particularly when the latter are weakly represented at the centre. Similarly, decentralisation can take different shapes. In 2002, virtually all the chiefs of police were from the province where they served, as were the large majority of the officers. In practice, therefore, policing was local and decentralised even if in theory the structure remained centralised. Foreign diplomats often thought that locally recruited police chiefs (as in 2002) could play a positive role because of their involvement in dispute resolution, even if there was no question that ‘local’ chiefs of police took the job because they were strongmen in control of large armed groups.59 On these points see International Crisis Group, Policing in Afghanistan: Still Searching for a Strategy (Brussels, 2008) and Andrew Wilder, Cops or Robbers? (Kabul: AREU, 2006). 57
See International Crisis Group, Policing (FN 57); interviews with Minister Jalali and Minister Atmar, Kabul, April 2011; interviews with senior police officers, Kabul, October, November 2010 and April 2011. 58
Personal communication with Shah Mahmud Miakhel, former Deputy Minister of Interior, July 2011. 59
Interview with foreign diplomat, Kabul, November 2010. A second effort occurred in 2006, during the tenure of Minister Zarar, largely under international pressure. This continued the efforts already gearing up in 2005. This reform effort had three main aspects: pay and rank reform, which introduced a new pay scale and cut down the originally hugely inflated number of general and other senior officers; an appointments board to vet appointments; and a trust fund (LOFTA) to mobilise international funds and handle paying policemen. In 2006 a probation board was established, in response to the failure of the appointments board to contain political interference. Pay and rank were reorganised and rationalised; LOFTA would prove particularly important in 2008 and onward in reducing corruption during the distribution of salaries.56 56
AAN Thematic Report 07/2011 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context The risks of locally recruited police forces were however obvious. In particular, local police forces became involved in local conflicts, at the expense of some of the communities. In Shindand district in 2007, for example, the police split along tribal lines, Barakzai versus Noorzai, and started fighting each other.60 Moreover, these police officers, locally recruited on the basis of patronage and local influence, proved often impossible to control. Finally, the interests of provincial strongmen and political factions might sometimes coincide with those of sections of the population, but not necessarily; even when they do, it is unlikely that all segments of the population feel similarly protected by a locally raised police force. Even in a decentralised system, to ensure that local police forces are not factionalised or biased against some communities, a strong centre is still necessary, to select and appoint, oversee and reward or punish. Without that, what you have is not decentralisation but fragmentation. Options for extending some control over the periphery were limited. From 2003 onwards, the MoI relied on a mix of local and semi‐local recruitment. The latter implied taking recruits from one district or province and sending them to a neighbouring district or province, so that they would be out of area but not too far from their families. For example, in 2007 most of the police of Garmser (Helmand) were from another district of the same province, Nawa.61 Because the salaries were too low to justify serving in a province other than their own,62 most police continued to serve in their own district or a neighbouring one. Still the reliance on out‐of‐area police as the main tool of re‐centralisation gained ground within the MoI for several years after 2002, as the central recruitment of officers and the establishment of a few mobile police units made headway. By the second half of the decade, in some areas a significant portion of the provincial police force was now loyal to the centre, rather than to local strongmen and factions. This was enabling for chiefs of police or governors who were inclined to improve law and order. In Helmand, for example, Governor Daud estimated in 2007 that of 1,700 policemen in the province, 250 were loyal to him and to the central government.63 Similarly, after 2004 Herat’s police force was divided between supporters of the 60
Interview with ex‐Deputy Minister Abdul Hadi Khalid, Kabul, October 2009. 61
Doug Beattie, An Ordinary Soldier (London: Simon & Schuster, 2008), 196. 62
Interview with Chief of police of Faryab, March 2009. 63
James Ferguson, A Million Bullets (London: Bantam Press, 2008), 154. central government and supporters of local strongman Ismail Khan, whereas until 2004 it had been entirely dominated by the former. However, in terms of providing security the gains for the local population were not obvious (see Section 6). By 2006–8 the debate among donors and within the MoI over centralised versus local police took a new shape. It occurred less and less between opponents and supporters of local strongmen; instead proponents of village‐based police forces were increasingly pitted against those who believed such forces could never be controlled by the centre. Although village‐based police could be expected to maintain links with strongmen, this represented a widening of the debate. Experiments in local policing succeeded each other (Auxiliary Police, Afghan Public Protection Force, Afghan Local Police), mostly confirming the fears of the pessimists.64 The potential of the police to collaborate with communities and local authorities was further reduced by lasting disagreement over who ultimately was in charge of the provincial police: the governor or the provincial chief of police. This has remained a point of contention between the MoI and the IDLG.65 In Jalali’s time, the effort to bring in educated recruits and shape them into a force responsive to MoI command took the shape of the Quick Reaction Force, which evolved into the Afghan National Civil Order Police (ANCOP) in Atmar’s time. Initially, the ANCOP was successfully deployed to temporarily replace police station crews taken for re‐training (the so‐called Focused District Development), leaving a positive impression among the population.66 After some initial success, however, by late 2009, this effort stalled because of the declining quality of recruits.67 Much of what was achieved in terms of 64
See Matthieu Lefèvre, Local Defence in Afghanistan: A Review of Government‐Backed Initiatives, AAN Thematic Report 03/2010 (Kabul/Berlin: Afghanistan Analysts Network, 27 May 2010). 65
Personal communication with Martine van Bijlert, Afghanistan Analysts Network, March 2011. 66
According to ISAF’s opinion poll surveys at least; anecdotal evidence suggests that indeed ANCOP had a better reputation than the police units it replaced. 67
Dion Nissenbaum, ‘U.S. Forces Pin Hopes on New Afghan Civil Police’, McClatchy Newspapers, 24 March 2010; Dion Nissenbaum, ‘Corruption, Incompetence Charges Plague New Afghan Police Force’, McClatchy Newspapers, 10 May 2010; Office of the Special Inspector General for Afghanistan Reconstruction, Actions Needed to Improve the Reliability of Afghan Security Force Assessments (Washington: SIGAR, 29 June November 2011 15
16 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
centralisation, therefore, was the result of expediency (injecting a number of non‐local policemen into the provinces, appointing a few chiefs of police loyal to Kabul here and there) rather than a coherent plan. The only element of planning to have an impact was the mobile force under direct control of Kabul, whose development initially faced much resistance.68 The Herat case study (Section 6) shows that this type of centralisation did not necessarily represent progress. 4.3 Combating corruption Establishing police units under the direct control of the MoI and, in doing so, increasing the leverage of the centre versus the periphery proved possible. However, the fight against corruption proved much more difficult to carry out – despite increased urgency and pervasiveness of corruption from 2001 onwards. Afghanistan ranked as one of the most corrupt countries in the world, and pressure from both the Afghan public and foreign donors to do something about it started mounting. Corruption has exacerbated the dysfunctionality of the Afghan policing system and negatively affected the morale of the police and the quality of the leadership; many mentors believed that corruption was sabotaging all their efforts to develop the police. From 2002 onwards, the misuse or diversion of resources by police personnel became common practices, reducing the effectiveness of an already weak system.69 Police corruption has reached a level of dysfunctionality where appointments to positions of power are disrupted and tasks routinely not fulfilled (some forms of corruption do not necessarily lead to dysfunctionality, for example charging unofficial taxes for services without disrupting their delivery.) Evidence of the police force’s involvement in corruption is overwhelming, even in official statistics. Until the first quarter of Afghan year 2009–10 (1388), 18,276 police personnel were accused of having been involved in 10,480 cases of misbehaviour (abuses of power and corruption), according to the attorney general’s office; 30 were 2010), 20; C.J. Chivers, ‘Afghan Police Earning Poor Grade for Marja Mission’, New York Times, 1 June 2010. 68
Interviews with former high level officials of the MoI, Kabul, April 2011. 69
Office of the Special Inspector General for Afghanistan Reconstruction, Actions Needed to Improve the Reliability of Afghan Security Force Assessments (Washington: SIGAR, 29 June 2010), 19–20. AAN Thematic Report 07/2011 generals.70 Having over 20 per cent of the police force implicated in reported cases of misconduct is very high, particularly since it is unlikely that authorities have identified all the cases. A survey carried out by UNAMA in 2007 found a negative service record (mostly related to criminal behaviour or corruption) for 38 per cent of the 2,464 police officers included in it.71 Mentors and trainers were aware of the extent of the problem.72 That extortion and misconduct continued on a large scale in 2010 indicates that only a minority of cases were being uncovered or followed up. Complaints by drivers and passengers concerning the behaviour of police at check‐posts on the highways remained very common.73 The only aspect where some success was achieved against corruption was in ‘ghost policing’ – where senior police officers claim salaries for non‐existing policemen and pocket them. The adoption of a new system of management and transfer of payment reduced the number of ghost police in the ranks, although donor sources believed that the actual number of police at the end of 2010 still fell significantly short of the theoretical payroll: perhaps by 30–40,000 out of 122,000 (compared to 36,000 out of 96,000 in autumn 2009).74 In 2010, attempts to assess the actual strength of police units showed absence rates of 20–25 per cent.75 NTM‐A (NATO Training Mission–Afghanistan) sources agreed that the problem of ghost police, although reduced, was still significant. New Interior Minister Bismillah Mohammadi started a tour of provincial inspections just after his appointment in 2010, to verify that the police were actually there and whether they were getting paid.76 A major problem in eliminating ghost police was that because of the force’s high attrition 70
‘Crime by Afghan Police Rises “Ten‐Fold” since 2002’, Kabul Weekly, 1 July 2009. The figures are however on uncertain reliability, as the Attorney General Office has not confirmed them to EUPOL. 71
Pratap Chatterjee, ‘Policing Afghanistan: Obama’s New Strategy’, CorpWatch, 23 March 2009. 72
See for example, Richard A. Oppel Jr., ‘Corruption Undercuts Hopes for Afghan Police’, The New York Times, 8 April 2009. 73
‘Drivers Face Police Bribes on New Road’, Tolo News, 19 June 2008; Soraya Sarhaddi Nelson, ‘Despite New Highways, Afghans Drive at Own Risk’, NPR, March 16, 2008; ‘No End to Police Corruption on Herat‐Kandahar Highway’, Pajhwok Afghan News, 1 September 2008. 74
Meeting with international police assistance officer, Kabul, 6 November 2010; meeting with NTM‐A officer, November 2010; interview with senior MoI official, Kabul, November 2010. 75
Interview with foreign diplomat, November 2010. 76
Interview with former senior MoI official, Kabul, October 2009. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context rate, determining whether chiefs of police reported attrition quickly was difficult.77 The negative national picture does not imply that some provinces were managed more effectively than others. The experience of Herat, detailed in Section 6, suggests that corruption can be contained even in the absence of professional officers, as long as a charismatic personality gave some kind of oversight and salaries were sufficiently high. In the presence of weak institutions, however, the best one can expect is a hit and miss pattern where in some provinces leaders are able (and willing) to enforce a relatively functional form of policing and in others they are not. In Afghanistan after 2001 the latter were the large majority. In principle, corruption has a negative impact on para‐militarisation because it represents the centre’s loss of capacity to control, particularly when corruption reaches the stage of dysfunctionality discussed above. A very corrupt paramilitary force is always going to be ineffective, even if such ineffectiveness is temporarily overshadowed by other weaknesses, such as poor leadership (i.e. the inability to come up with an operational plan and to mobilise resources effectively). 4.4 The rule of law The main concerns of donors often diverged not just from those of the provincial strongmen, but also from those of the reformers within the Afghan cabinet. Sometimes convergences occurred, for example on the issue of reducing political patronage to the benefit of more meritocratic criteria. Sometimes donors disagreed among themselves. The American military, for instance, was mostly concerned with increasing the capabilities of the police force to suppress the insurgency, whereas the Europeans mostly worried about the weakness of the rule of law in Afghanistan and wanted the MoI to improve it. The Afghan reformers had mostly in mind the strengthening of the Afghan state. With regard to the implementation of the rule of law, even in the second half of 2009, UNAMA reported that detention beyond the custody limit and on insufficient or no grounds by police chiefs was widespread, particularly in northern Afghanistan where ISAF mentoring was limited. Chiefs of police seemed to believe that the 72‐hour 77
Interview with foreign diplomat, November 2010. custody limit was discretional.78 Administration was limited, as demonstrated in the case of the big Kandahar prisoner escape of 2008, when it transpired that none of the 900 inmates had been fingerprinted or photographed.79 Searches remained poorly organised and the habit of arresting all bystanders in the event of a crime was still common.80 Apparently, few in the MoI considered strengthening the rule of law a priority, given the state of war in the country and the limited range of appropriate human resources.81 The commitment of energy and resources to establishing the rule of law is certainly not what is slowing down effective para‐militarisation in Afghanistan. While paramilitary policing is not incompatible with the rule of law, in a context of ongoing war – unsurprisingly – the form of para‐militarisation that is asserting itself is not very sensitive to the rule of law. In the case of Afghanistan, the absence of strong political pressure in support of the rule of law has made this disregard particularly pronounced. 4.5 What has been achieved? Initially international investment in the police was modest, as the lead nation (Germany) was not willing to spend billions of dollars. Indeed there is little evidence that any of the Western powers involved in Afghanistan advocated a major effort focused on the police. In 2005, the American role started expanding, with money and consultants flowing to the MoI. The real escalation of American support started in 2008, when billions were being committed. By 2011, the international community had spent over US$15 billion on Afghanistan’s police; a few hundred advisors or trainers and (eventually) over 2,000 mentors were assisting the development of the police. However, it is far from clear that the massive resources always had a positive impact; the huge cash flow created an 78
‘Many Afghans Being Detained Unlawfully – UNAMA’, Pajhwok Afghan News, November 2009. 79
International Crisis Group, Reforming, 7 (see FN 57). 80
Nadene Ghouri, ‘Afghan Police: Corrupt and Brutal, and Still Not Fit for Purpose’, The Guardian, 6 November 2009. 81
For a more thorough discussion of the rule‐of‐law issue, see International Crisis Group, Reforming (FN 57); Laurel Miller and Robert Perito, Establishing the Rule of Law in Afghanistan (Washington: USIP, 2004); United States Department of State and the Broadcasting Board of Governors Office of Inspector General, Rule‐of‐Law Programs in Afghanistan, Report Number ISP‐I‐08‐09 (Washington, January 2008). November 2011 17
18 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
environment where too many competed for the favours of the expatriates distributing the resources, disrupting coherent policy making.82 The reasons for the less‐than‐modest results, as illustrated in the previous sections, are several. Education and professional training were neglected; recruiting junior officers of adequate level was a problem. Recent estimates put 11 per cent of patrolmen and 35 per cent of NCOs (non‐
commissioned officers) as literate.83 The implications of a largely illiterate police force are obvious, among others hampering familiarity with the laws of the country. Paramilitary policing was also negatively affected and police officers were reportedly unable to read maps. The low educational level of the police force made it difficult to enforce record‐keeping and reporting. Eventually the MoI, perhaps under pressure from NTM‐A, started developing its own strategy for retaining qualified officers, introducing rules to prevent private security companies from poaching the best police cadres, mainly through tighter contracts that made quitting before their duration illegal (see also Section 5.4).84 At the time of writing there were no clear signs that this strategy was working, as available retention data did not discriminate between officers and patrolmen. By 2010, continuity arguably prevailed over change. With the partial exception of ghost policing, corruption was as prevalent as ever, while appointments were still largely driven by political consideration or by sheer corruption (see 4.3 above). On the positive side, the MoI had gained some leverage over its provincial units, but a few thousand ANCOP policemen used as a central reserve were still too few to have a major impact on the rest of the police. ANCOP’s ranks were continuously expanded, but with a dilution of the quality of the recruits and a resulting diminution of its reliability (see 4.2 above). The new development in 2010 was the decision (agreed between the international community and the Afghan government) to proceed towards ‘transition’ and the gradual handing over of responsibility to the Afghan security forces. As the Afghan police and army began to take over, and confronting a determined and resourceful armed opposition, the solidity of the MoI command and 82
Personal communication with Shah Mahmud Miakhel, July 2011. 83
Katharine Houreld, ‘Afghan Police Learn to Shoot, Patrol – and Read’, Associated Press, 22 November 2010. 84
Matthew Green, ‘Afghan Drive to Stop Poaching of Police Recruits’, Financial Times, 25 January 2010. AAN Thematic Report 07/2011 control structure became a much greater concern than ever before. In the increasingly demanding environment which was emerging, would the MoI be able to control a police force largely built on the basis of political patronage and riddled with corruption, that was now being handed greatly expanded responsibilities? To the extent that anything was achieved, state security against the insurgency improved (as discussed in Section 5 below). The rule of law and the security needs of the population still ranked very low among the MoI’s priorities. A clear model of policing was still missing: a gendarmerie (ANCOP) had been established and was being expanded, the Criminal Investigation Department was slowly being developed as a civilian police force, the Border Police was slowly being upgraded into another gendarmerie‐type force, but the Uniformed Police, which represented the bulk of the police and in particular the provincial units discussed throughout this chapter, was still more like a fragmented coterie of militias than either a paramilitary police or a civilian police force, and was moving in neither direction (see Section 5 on this point). 5. CONTINUITY AND CHANGE IN COMMAND AND CONTROL Both in Washington and in Kabul the clear sense in 2011 was that the priority was paramilitary policing, and the training imparted to the Uniformed Police was mostly paramilitary. Why then were the achievements so modest? Paramilitary forces are not defined as such primarily because of the training they receive, but because of the organisation and structure they have. This section is dedicated to the command and control structure of Afghanistan’s police and seeks to determine whether the police force is ready to face the paramilitary tasks which lie ahead. The section discusses the various aspects of command and control in relation to the viability of paramilitary policing in Afghanistan, relating not only to ANCOP and ABP but also to AUP, which does not describe itself as a paramilitary force but is deployed and trained as if it were. The chain of command is specifically discussed in Section 5.1, while Section 5.2 reviews a particular aspect of control: reporting from the bottom to the top, (as a leadership that does not know what is going on at the bottom of its organisation cannot take decisions in an effective way). Other aspects of Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context control relating to various facets of indiscipline and the efforts to address them are discussed in Sections 5.3, 5.4 and 5.5. Although the structures and institutions to maintain discipline within Afghanistan’s police force exist, they are not very effective. The last section (5.6) discusses whether the command and control model adopted in Afghanistan is most suitable for the human resources available and for the social structure of the country. 5.1 The command structure The quality of command and control is an essential feature of paramilitary policing, more so in the context of a fragmented and politically divided society like Afghanistan’s. Command and control is what gives the police the capacity to follow a strategy and act in a coherent way when facing the insurgency. In the early years of the post‐Taleban era, centralised command and control was virtually non‐existent. A number of individuals had influence over portions of the police force, while institutions were absent, as found by Amnesty International Researchers in 2003: At least five senior ranking officials held positions apparently with responsibility for policing. General Assefi was appointed as General Commander of the National Police in September 2002. Despite his title, he appears to have the least control over the police. When interviewed by Amnesty International, he admitted that he does not have authority over the police, who are loyal to local commanders rather than to central government. General Hellal, a former air force pilot, is Deputy Minister for Internal Security. He told Amnesty International that provincial commanders report to him regularly. General Salangi is the head of Kabul police, has a ten thousand strong force and holds much actual power. He has a military background. General Naseri, Special Adviser to the Minister of the Interior, also appeared to have some responsibility for policing. General Jurat, the ‘Head of Security and Public Controls’ also holds a powerful role in Kabul, controlling his own force of approximately 4,000 police. They patrol Kabul 24 hours a day, and control checkpoints at the airport and at the gates of the city. His police force operates separately from the Kabul police force that is under the command of General Salangi. General Jurat told Amnesty International that he reports directly to the Minister of Interior.85 In some cases the command‐and‐control structure was so weak that knowing who was the chief of police in charge at a given moment was difficult. In one case in Kandahar in 2003, the Americans had to demand that Khan Mohammed, who was claiming the post of chief of police, produce a letter of appointment or be forcefully removed.86 Throughout the period being discussed (2001–11), the chiefs of police often did not know exactly how many men were effectively under their command.87 Various factors contributed to make the MoI’s chain of command extremely ‘complicated’:88 



the initial Turkish imprint of centralisation, dating back to the 1920s; the heritage of the Soviet model of decision making, but without a ruling party to function as a source of oversight; the predominant attitudes towards authority, privileging personal relations over institutional frameworks; and the political realities of the day, with much political interference being exercised by different individuals and groups. The extreme centralisation (in principle) of the command‐and‐control structure meant that units had no right to take decisions; because of this no effort was done to develop their analytical capacity. The problem was compounded by the limited capacity of the MoI HQ to process information and take quick decisions. Among other things, in 2002–3 the MoI had little or no capacity for strategic thinking.89 The situation improved only marginally after that. In reality, throughout the 2001–6 period, the police in the provinces commonly resisted orders from Kabul.90 The gradual improvement of the technological resources strengthened the role of provincial HQs, at least enabling senior police officers to work as a team if they wanted to; the 85
Afghanistan: Police Reconstruction Essential for the Protection of Human Rights, AI Index: ASA 11/003/2003 (London: Amnesty International, March 2003), 8. 86
UNAMA internal source, 20 May 2003. 87
For a case in then relatively quiet Chahardarah see Candace Rondeaux, ‘A Ragtag Pursuit of the Taliban U.S. Effort to Train Afghans as Counterinsurgency Force Is Far From Finished’, Washington Post, 6 August 2008. 88
NTM‐A, ‘90 Day Assessment’, 1 October 2010. 89
Interview with foreign diplomat, Kabul, November 2010. 90
Interview with MoI official, May 2006, Kabul. November 2011 19
20 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
problem was often the lack of capacity and sometimes the political will among senior officers to implement decisions taken in Kabul.91 During the second phase of MoI reform, which started in 2006 under Minister Zarar, some small but important steps were taken to ensure that the centre had some control over the force. The improving quality and the updating of MoI personnel databases (non‐existent in 2002) allowed it to plan some improvements in its administration.92 In 2006 a new National Police Command Centre was opened, with advanced communications facilities that enabled the MoI to communicate with its provincial units, the Army command‐and‐control centre and units in the regions. A secure system of communication was established, featuring videoconferencing capabilities at the zone level. Reports from the districts were delivered via mobile phone messages and radio communication with the provincial HQ; this was standard by 2005.93 But although the technical capacity of the National Command Centre was greatly improved, the large international presence there and in the regional centres meant that its ability to function autonomously in an integrated way remains questionable at the time of writing this report.94 New technologies notwithstanding, as late as November 2010, bypassing the chain of command was still a frequent occurrence when officers had political connections (that is, quite often).95 The UN rapporteur wrote that he received reports of cases in which police did attempt to carry out investigations, but senior government officials interfered with or prevented the investigations.96 Shifts were not organised in a bureaucratic way, but were based on personal connections.97 Even in 91
Libra Advisory Group, briefing at King’s College London, 2009. 92
‘Afghanistan Security: U.S. Programs to Further Reform Ministry of Interior and National Police Challenged by Lack of Military Personnel and Afghan Cooperation’, GAO‐09‐280 (Washington: GAO, March 2009), 25. 93
Interview with Haidar Basir, former deputy minister, Kabul, October 2010. 94
Interview with foreign diplomat, Kabul, November 2010; ‘U.S‐funded National Police Command Center Opens Dec. 16’, Combined Forces Command – Afghanistan, 17 December 2006. 95
Interview with senior police officer, 2 November 2010. 96
‘Report of the Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions, Philip Alston. Addendum: Mission to Afghanistan’, General Assembly, A/HRC/11/2/Add.4, 6 May 2009. 97
Meeting with NTM‐A official, November 2010. AAN Thematic Report 07/2011 2010, the MoI had huge problems simply disseminating information from the top to the bottom; chiefs of police were often unaware of recent decisions and developments.98 When messages from the top were communicated to the units on the ground, implementation still varied according to local context. Sometimes orders were implemented: in Herat for example the directive to prosecute bad police drivers resulted in police driving getting better.99 Often they were not. One example of disobedience of a direct order by a police officer on the ground occurred during the 2010 elections: during a demonstration by defeated candidates from Baghlan province, the police were ordered not to let the demonstrators into the Electoral Commission’s compound. The police commander, who was a relative of one of the candidates, decided instead to let them in without even searching them.100 In general, NTM‐A and ISAF described the MoI system as ‘not joined up’,101 and the command and control of the police as ‘at best fragile’ and ‘generally reliant on key commanders’.102 Where signs of dynamism are evident at the provincial level, they might not be due to an increase in support from the MoI, but to the appointment of a charismatic and capable chief of police. Some good managers existed in the system as of 2010 and under an effective chief of police, performance could improve; there was some willingness to get the police force’s act together at least in terms of fighting back the insurgents and strengthening discipline among the hard core of police professionals formed in the 1970s and 1980s.103 The more proactive chiefs of police would travel to the districts (admittedly with ISAF facilitation), inspect their men and assess their needs, although in several provinces such activities were made difficult or risky because of the insurgency.104 Although professional officers (that is, officers educated as police) were tangentially more disciplined than the rest and therefore more dependable for the MoI, they were in short supply, 98
Meeting with NTM‐A official, November 2010. Meeting with Afghanistan Public Policy Research Organisation (APPRO) analyst, Kabul, October 2010. 100
Interview with foreign adviser, Kabul, November 2010. 101
Meeting with NTM‐A officials, November 2010. 102
‘Afghan Police Complete Command, Control Exercise’, NATO News Release, 30 November 2010. 103
Libra briefing (see FN 92). 104
Meeting with Gen Azizi, chief of police of Paktia, London, June 2009. 99
21
Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context not always keen to be deeply involved in the counter‐insurgency effort and sometimes politically suspect (having a leftist past). Appointing provincial and zone chiefs of police personally or politically loyal to the minister emerged as the overriding strategy to allow the ‘moderate meritocracy’, as he was considering the capabilities of his subordinates in making appointments. The system, after all, was built to respond to strong leadership; the expectation of Map 1. Professional and political background of provincial chiefs of police, April 2011 Sources: Interviews with MoI and UN officials, as well as with security and political analysis based in Afghanistan, 2009–11. MoI to exercise command in the provinces. Map 1 shows the picture as of April 2011. The bulk of the provincial chiefs of police were either professional officers, in part appointed under international pressure, or personally or factionally (J‐J in the map, referring to jihadis who were specifically members of Jamiat‐i Islami, Mohammadi’s party) linked to the minister of interior (Bismillah Mohammadi); the few who were not had been appointed under his predecessor. the patrolmen was to be led.105 The main limitation of his effort, however, was that the system did not have the capacity to identify, select and promote new talent from the ranks, forcing the minister to rely on a small and shrinking pool of capable and loyal people (according to his own assessment) to make appointments. The MoI did not train a new generation of paramilitary police commanders either: the police academy was focused on civilian policing and the paramilitary training provided by either the military or by private contractors was extremely basic. The number of charismatic police leaders declined due to Taleban targeting; during 2011 the most notable casualties were the head of the Northern Zone, Daud Daud, and three prominent provincial chiefs. With the killing in 2011 of Khan Mohammed, chief of police of Kandahar, Minister Mohammadi run out of credible loyalists to run this key province and What a proactive minister like Mohammadi had to do to instil some commitment (to counter‐
insurgency if anything else) and energy into the police force was to appoint to key positions people who were loyal or connected to him and ask them to commit themselves to the fight. While least some of the people Mohammadi chose for key jobs were undoubtedly capable fighters, they were all former comrades in arms of the minister during the 1980s and 1990s. In this sense, what Mohammadi was striving for was his own kind of 105
Interview of chief of police of Maimana city, November 2004. November 2011 22 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
eventually had to appoint Colonel Razzeq to the post, who was linked to both the Karzai family and Gul Agha Shirzai, another strongman opposed to Mohammadi’s political faction. If having a charismatic and professional chief of police at the top of the provincial structure was still the decisive factor in 2010 in the performance of the police, even when such a person was present, command and control at the tactical level was precarious at best; while logistics and planning were always weak, even in the best provinces.106 Until a critical mass of professional officers was created, a single, professional chief of police at the provincial level could only go so far, often being unable to impose discipline even on his chiefs of police in the city districts if he was not particularly charismatic or determined.107 At the central level, specific departments showed signs of growing maturity by 2010–11: decisions about big operations were processed relatively quickly, although political interference and lack of communication were still an obstacle. In the case of the 2010 Nangarhar counter‐narcotics raid which involved Russian agents, the ministry cleared it remarkably quickly, even if President Karzai then protested that it violated the sovereignty of Afghanistan.108 The technological capacity to communicate with the provinces improved immensely, although not always put to good use. Where the MoI performed worst was in organising and coordinating rapid responses to crises, a consequence of weak management. The MoI of 2010 could implement a plan, but not react quickly to contingencies. For example, when riots occurred in October 2010 in Kabul between Hazaras and Kuchis, the police could not cope and the minister himself had to deploy on the ground to manage the situation.109 In sum, the main progress in developing a solid chain of command within the MoI after 2001 pertained to the realm of technology, which improved dramatically. In terms of institutional development, however, little change occurred between 2002 and 2011. At times, particular personalities within the MoI proved more effective than their colleagues and improved the security of 106
Interview with foreign diplomat, November 2010. Interview with police officer, Kandahar, January 2006. 108
Interview with foreign diplomat, November 2010; Laura King, ‘Karzai Denounces Drug Raid in Afghanistan’, Los Angeles Times, 31 October 2010. 109
Interview with foreign adviser, Kabul, November 2010. 107
AAN Thematic Report 07/2011 one particular province or region, but with little or no lasting effects. As soon as they were removed, the police reverted to the dysfunctional state found in most of the country. 5.2 Reporting to the centre Record keeping and reporting to the top might be boring features of modern bureaucracies, but they limit room for arbitrary behaviour at the bottom. In 2002–5 a series of visits to police stations around Afghanistan showed that orders and communications were being issued orally or written on plain paper and copies were not being kept.110 Other surveys came to a similar conclusion: ‘Records of arrests and incidents, or other logs, are not being kept in any of the police stations that Amnesty International visited in Afghanistan.’111 In 2003 the police often did not even have paper or pens.112 Whatever paperwork and written communications were being exchanged between police officers was very loose and disorganised in the early post‐2001 days; for example, a letter by Nangarhar Chief of Police Hazrat Ali (2003–4) was on the letterhead of the Bank of Afghanistan and signed by Hazrat Ali, who addressed it to ‘All Security Guards and Policemen’, advising that an individual was missing, and instructing them to help ‘find the person he suspects’. It featured no names, no addresses and no explicit order to investigate the murder alleged by the complaining individual.113 In the following years some improvement in reporting back to Kabul and in record keeping took place, sometimes facing serious resistance within the MoI. In the words of a former MoI official, Minister Atmar (2008–10) asked police stations to keep files, but that did not make him popular.114 By 2009, some police stations were keeping a rudimentary record of prisoners, but little effort had been made to create a proper administrative system; procedures hardly existed.115 From 2002 onwards, the MoI considered reporting on counter‐terrorism as deserving priority over other forms of reporting. Using protected mobile phones, reports about incidents in the provinces started being relayed to the MoI and were filed 110
Personal observation. Afghanistan: Police Reconstruction Essential, 28 (see FN 86). 112
Personal observation. 113
Paul Watson, ‘Afghan Gangs on Rise’, Los Angeles Times, 21 May 2005. 114
Interview with former senior MoI official, Kabul, October 2010. 115
Interview with foreign adviser, Kabul, October 2009. 111
Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context there. Reporting received a potential improvement in 2009–10 with the introduction of the Afghan Police Incident Reporting System (APIRS), a standardized reporting system in which incidents are consolidated in a national database and automatically analyzed. The program generates a multitude of reports: logistics, intelligence, readiness, casualty and operations. … Presently, there is no consistent reporting database in place. Reports are hand written and then saved as text documents. This often results in unreliable information, inhibiting operations and intelligence collection efforts. “Sometimes in a situation we don’t even know how many personnel we have available,” said Colonel Nematullah Haidary, Operations General Director of the ANP. … The driving forces behind its use in Afghanistan are the police mentors of CSTC‐A.116 As of May 2011, however, among the MoI field units, only Police District 3 of Kabul city was able to use APIRS.117 If the technology and the technique of reporting undoubtedly improved, the willingness of the field units to report correctly remained in doubt. In the words of a MoI adviser, ‘nobody wants bad news and there is a lot of misreporting up the chain of command.’118 Information rarely travelled upwards, even in 2010.119 Statistical data was transferred to Kabul slowly, typically a year late. Moreover, statistics were not reliable. As a result, the MoI did not really know in detail what was going on. The MoI was not seen as a ‘mother organisation’. In practice, more information existed at the provincial level than at the centre.120 A reason for the frequent failure to file incidents was the desire not to alienate communities and strongmen, particularly those better‐connected with the government. Therefore, only a portion of crimes and incidents was being recorded. This may have been even worse than not recording crimes at all because it alienated the communities which were less connected to government officials.121 This was a particularly strong problem in the south, where the strongmen and the communities most closely connected to Kabul were in a position to get sympathetic chiefs of police appointed.122 The other weakness in the reporting system was the absence of a mechanism in the MoI to consolidate the data received. The National Police Command Centre in 2010 still used a spreadsheet to capture what was going on, with no ability to analyse trends and compile statistics; the capacity was limited to basic quantitative assessments.123 What the MoI could then issue were daily incident reports, detailed but of dubious quality. Certainly they often missed incidents, but whether this was due to deliberate misreporting or simply due to the police not being aware is difficult to say.124 Improvements in the reporting system of the MoI were significant after 2001, but on the whole reporting remained very weak, preventing the MoI leadership from having a clear picture of the situation on the ground. The ambition to create a complex reporting system, at least in the short and medium term, produced an extremely inefficient organisation, where centralisation did not serve the purpose of coordinating different components of the MoI, but only slowed everything down. From a certain perspective, inefficient record keeping made Afghan citizens worse off than if paperwork had not existed, as the MoI was not able to manage the paperwork it produced. The system was very centralised and the middle ranks did not want to take any responsibility, so in the words of one adviser to the MoI, ‘everything floats up to minister and deputies, but there is a bottleneck there’.125 Records cluttered the system, slowing it down, with few people bothering to check the paper trail to ensure accountability. 5.3 Control: Collaboration with the enemy The worst cases of misconduct, at least from a state’s perspective, include collaboration with the insurgents. Cases of police selling weapons and ammunition on the black market or even directly to the insurgents are common; one episode of police handing over weapons to the Taleban was even filmed and shown on a British Broadcasting 116
‘ANP Tests National Reporting System’, Enduring Ledger, June 2009. 117
EUPOL, Newsletter of the European Union Police Mission in Afghanistan, 19 May 2011. 118
Meeting with foreign adviser, September 2009. 119
Meeting with APPRO analyst (see FN 100). 120
Ibid. 121
Interview with former deputy minister Abdul Hadi Khalid, Kabul, October 2010. 122
Interview with senior MoI official, Kabul, November 2010. 123
Interview with foreign adviser, Kabul, November 2010. 124
Interview with EUPOL officer, Kabul, November 2010. 125
Interview with foreign adviser, Kabul, September 2009. November 2011 23
24 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
Corporation documentary.126 Perhaps the most serious episode of policemen allegedly collaborating with the insurgents was the April 2008 attack on President Karzai, who was presiding over a military parade in Kabul.127 There were also indications that MoI employees, among others, facilitated the attack on the Kabul Serena hotel in January 2008.128 Often eyewitnesses have reported instances of police collaboration with the Taleban, or at least passive tolerance of their activities.129 Reports of police helping the Taleban to carry out attacks against US, Danish or British units have also surfaced.130 Proactive police officers sometimes allege that they are isolated in their struggle with the insurgents and are singled out for attacks, while surrounding police stations are left alone.131 Apart from the attack on Karzai, all reported cases of Taleban infiltration were at the lowest level of the MoI hierarchy.132 However, mentors and ISAF believed that several operations were compromised because of Taleban infiltration of the police; in Helmand in 2009 the police were kept out of planning to avoid forewarning the Taleban.133 According to UN sources, many chiefs of police in the districts of the south and southeast had contacts with the Taleban and developed agreements with them.134 Diplomatic sources highlighted that even in the north, deal making between police and Taleban might be a problem, for example in the case of Almar in Faryab.135 At the MoI, high ranking officers admit that in some areas there might be informal ceasefires with the 126
‘Afghan Ministry to Probe Footage Showing Police Handing Over Weapons to Taleban’, Aina TV, 26 September 2009. 127
Carlotta Gall and Waheed Wafa, ‘Plot to Kill Afghan Leader Was Inside Job’, New York Times, 1 May 2008. 128
Joanna Wright, ‘Changing the Guard – Efforts Continue to Resurrect Afghan Police Force’, Jane’s Intelligence Review, 16 October 2008. 129
‘Police Let Taleban Take Us Away: Freed Afghan’, AFP (Agence France‐Presse), 10 December 2007. 130
Eric Schmitt, ‘Afghan Officials Aided an Attack on U.S. Soldiers’, New York Times, 3 November 2008; Stuart Tootal, Danger Close (London: Murray, 2009), 106; Patrick Bishop, 3 Para (London: Harper press 2007), 238. 131
Aryn Baker, ‘Policing Afghanistan’, Time, October 21, 2008. 132
‘At the Center of the Storm: An Interview with Afghanistan’s Lieutenant General Hadi Khalid – Part Two’, Terrorism Monitor 7 Issue 28 (17 September 2009). 133
Libra briefing (see FN 92). 134
Interview with UN official, Kabul, October 2010. 135
Interview with foreign diplomat, November 2011. AAN Thematic Report 07/2011 armed opposition.136 According to the chief of staff of the MoI, officers have already been removed for cooperating with the enemy.137 Although no overall figures exist, defections from the police appear to be frequent, particularly in Ghazni, Helmand and Badghis, but also Zabul, Nimruz, Farah, Herat and Wardak.138 In December 2010, the MoI spokesman reported that desertions from the police were on the rise, particularly when units were deployed to the areas most affected from the insurgency; he did not specify how many were joining the Taleban.139 In other instances, the police seemed to be cooperating with the Taleban on a commercial basis. Wardak police, for example, lost 160 weapons to Taleban in alleged clashes during 2005–6 and the suspicion was that they sold them for cash.140 5.4 Control: Indiscipline in the ranks In 2002–3, even the most basic features of discipline were missing from Afghanistan’s police, despite its paramilitary character. Common violations included not wearing a uniform, displaying symbols of factional affiliation, insubordination and violation of recruitment rules. The police commonly did not wear uniforms, even 136
Interview with senior MoI official, Kabul, November 2010. 137
Interview with General Nuristani, MoI head of personnel, Kabul, October 2007. 138
‘Forty Afghan Police Flee Checkpoints with Weapons after Not Paid in Full’, Pajhwok News Agency, 30 March 2006; Carlotta Gall, ‘Troops in Afghan District Find Anger at Lax Government’, New York Times, 18 September 2006; ‘5 Policemen Killed by Colleagues in S Afghanistan’, Xinhua, 26 November 2008; Julius Cavendish, ‘Afghan Police Unit Defects after Cutting Deal with the Taliban’, The Independent, 3 November 2010; ‘Commander among Three Policemen Slain by Colleagues’, Pajhwok Afghan News, 1 September 2007; ‘Cops Join Forces with Taliban’, Pajhwok Afghan News, 25 January 2010; ‘12 Afghan Police Involved in Robbery Surrender to Taliban’, Xinhua, 3 August 2008; ‘5 Cops Killed in Internecine Clash; 4 Kidnapped’, Pajhwok Afghan News, 30 July 2007; ‘Security Round Up: Afghan Police Join the Taliban’, Tolo News, 30 December 2008; ‘Dozens of Afghan Policemen Arrested for Helping Taliban’, DPA, 30 December 2008; Rod Nordland, ‘25 Afghan Police May Have Joined Taliban’, The New York Times, 18 February 2010; ‘Defections Hit Afghan Forces’, Aljazeera.Net, 15 October 2008; ‘Police Officer Kills 5 Fellow Police in Afghan East’, Aina TV, 15 June 2010; Tootal, Danger Close (see FN 131), 205; Sean Rayment, Into the Killing Zone (London: Constable, 2008), 95. 139
Tolo TV, 12 December 2010. 140
Interview with government official, Kabul, March 2009. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context in the cities.141 Even in subsequent years the resistance to using uniforms was strong, and weak commanding officers would not enforce the regulations.142 Often the police refused to wear uniforms for fear of reprisals from the insurgents, particularly when unaccompanied by ISAF units.143 Another common example of basic indiscipline was the display of symbols of factional affiliation on police cars or in offices. The most classical example is that Commander Massud’s pictures are displayed by police in Kabul and in the northeast. Initially this was tolerated by the MoI, itself badly factionalised, but when a ministry keen to show at least some outer signs of de‐factionalisation ordered the removal of these symbols, success was partial. Orders have to be repeated periodically and pictures of factional leaders remain prevalent.144 More‐serious forms of indiscipline have been common too. Police have refused to deploy for operation, particularly in the most dangerous areas.145 Sending them to do the holding job after an ANA/ISAF clearing operation proved to be the least popular destination.146 Absence from duty was a widespread problem, estimated by then‐
Minister Atmar at 20 per cent in 2008.147 Squabbles among policemen in some cases led to armed clashes and loss of life: for instance a dozen policemen were killed in a series of clashes among police groups in Kandahar province in 2007–8.148 Causes of the low level of discipline were several. Lack of equipment and supplies, in particular of fresh food, were certainly affecting morale negatively.149 Moreover, living conditions in the police stations were very poor, with tens of policemen sharing one or two room for a month before getting a meagre two days off.150 Even food allowances were inadequate, particularly in years of rising food prices.151 The misbehaviour of the police was also often linked to poor pay, delayed salary payments, or the embezzlement of salaries by somebody in the MoI.152 In fact, commanding officers were often forced to operate and live in constant fear of their men mutinying or deserting. The situation was compounded by the non‐existing logistics, which often left the policemen poorly equipped and without winter clothes, or completely reliant on American support.153 At one point some units of mentors started paying the police units directly to bypass the MoI, avoid delays and reduce corruption.154 From 2008 onwards, a major effort has been made through LOFTA to improve the reliability of the payment system (see also Section 4). Until then, the system entirely relied on the commanding officer to transfer the money to his subordinates, if the money was not embezzled before leaving MoI’s HQ. It was calculated that on average the policemen would only get 60–70 per cent of their meagre salary.155 MoI officials reportedly encouraged police officers in the field to lie to their men about the starting salary they were entitled to, to avoid having to justify payment shortfalls later.156 The absence of a system of shifts contributed to weaken morale and therefore discipline; those deployed to the worst areas of the country had little or no prospect of getting out. As one NTM‐A 141
‘Police Told to Remove Jihadi Photos from Vehicles’, Tolo News, 3 April 2008; Wali Arian, ‘Police Must Remain Neutral in the Polls: MoI’, Tolo News, 29 June 2009; Afghanistan: Police Reconstruction Essential (see FN 86). 142
Interview with senior police officer, November 2010; James Palmer, ‘Weak Afghan Police Threaten NATO Plan’, The Washington Times, 28 August 2008. 143
Rayment, Killing Zone, 92 (see FN 139); Bishop, 3 Para, 211 (see FN 131); Ann Marlowe, ‘Policing Afghanistan: Too Few Good Men and Too Many Bad Ones Make for a Grueling, Uphill Struggle’, Weekly Standard 014, Issue 14 (22 December 2008). 144
‘Police Told’, (see FN 142); Arian, ‘Police Must Remain’, (see FN 142); Afghanistan: Police Reconstruction Essential (see FN 86); personal observation. 145
Bishop, 3 Para, 211 (see FN 131). 146
Interview with foreign diplomat, November 2010. 147
International Crisis Group, Reforming, 2–3 (see FN 57). 148
Graeme Smith, ‘Kandahar Police Shoot It Out with New Foe – Themselves’, The Globe and Mail, 6 February 2008. 149
Meeting with NTM‐A officers, November 2010. Najib Khelwatgar, ‘Afghan Enforcement Officers Struggle with Poverty, Few Weapons, Little Ammo’, Canada Press, 18 April 2006. 151
Marlowe, Policing Afghanistan (see FN 144). 152
Habibul Rehman Ibrahimi, ‘Police Extortion Fears’, Afghan Recovery Report No. 32 (18 October 2002); Owais Tohid, ‘Cops Go Crooked in Kabul as Pay and Training Lag’, Christian Science Monitor, 3 July 2003. 153
C. J. Chivers, ‘Afghan Police Struggle to Work a Rough Beat’, The New York Times, 13 January 2008; Marlowe, Policing Afghanistan (see FN 144). 154
‘Afghan Police Salaries Paid by Canadian Military’, CBC News, 9 October 2007. 155
Bill Graveland, ‘Bank Cards Helping Afghan Police Slowly Shed Corruption Tag’, The Canadian Press, 19 November 2008. 156
Jason Motlagh, ‘Afghan Police Fight to Survive’, Asia Times, 28 August 2007. 150
November 2011 25
26 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
senior officer put it, ‘the shifts system can give hope to those serving in the mud’.157 Another factor feeding indiscipline was the insufficient number of capable and determined commanding officers, as discussed above. Trainees from the police academy in Kabul were openly stating their fear of being deployed in areas affected by the insurgency.158 Mostly the police showed little motivation to fight the insurgents, at least in the perception of NATO personnel.159 When the police did show motivation, it was rarely out of loyalty to the government but out of personal desires for revenge.160 Colonel Razzeq of the Border Police in Spin Boldak, famous for his lust for fighting the Taleban, had family members killed by Taleban.161 In fact people quite commonly joined the police in the south to obtain protection against the Taleban.162 Drug addiction was clearly another major source of police indiscipline and misconduct. During 2010, the MoI started admitting that drug addiction was a serious problem among the patrolmen, but still tended to underestimate its dimensions. Official figures indicated the existence of 1,800 addicts, following the testing of 114,000 policemen.163 Anecdotal evidence suggests that the level of addiction was much higher, with peaks in areas where opium derivatives or hashish are widely available. Partial testing carried out in 2008–9 found that no less than 16 per cent and up to 95 per cent of recruits were testing positive to drugs, depending on the sample.164 Attempts to 157
Interview with General Burgio NTM‐A, Kabul, 26 October 2010. 158
Jason Motlagh, ‘New Afghan Police Terrorized by Taliban’, Washington Times, 10 August 2007. 159
Nir Rosen, ‘Something from Nothing: U.S. Strategy in Afghanistan’, Boston Review, January/February 2010. 160
See, for example, Police Chief Khodaidad in Bala Buluk (Farah), as reported in Bilal Sarwary, ‘Bribery Rules on Afghan Roads’, BBC News, 29 July 2008. 161
Matthieu Aitkins, ‘The Master of Spin Boldak’, Harper’s Magazine, December 2009. 162
Interview with former Deputy Minister Abdul Hadi Khalid, Kabul, 26 October 2010. 163
‘Ministry Tackling Drug Addiction in Afghan Police’, National Afghanistan TV, 5 September 2010. 164
‘Need for Policing Charas Use by Cops’, Pajhwok Afghan News, 24 March 2006; Kim Barker, ‘Drug Abuse an Increasing Problem among Afghan Police, Soldiers’, Chicago Tribune, 5 October 2008; Martin Patience, ‘Drug Abuse Hampers Afghan Police’, BBC News, 18 February 2009; Per Anders Johansen and Tor Arne Andreassen, ‘Police Face Huge Training Challenges in Afghanistan’, Aftenposten, 29 September 2008; Office Of The Special Inspector General For Afghanistan Reconstruction, see FN 70, 20; Ann Marlowe, Policing, see FN 150. AAN Thematic Report 07/2011 systematically test the police for drugs faced resistance among the policemen. Rosen writes of one of these tests being administered in Helmand: only 53 of 80 policemen showed up; some refused to take it and 20 tested positive.165 Over the years, the MoI took some initiatives to improve discipline within the ranks. A claimed achievement in this regard was the distribution of ID cards and uniform badges to all police around the country; it struggled to distribute the police IDs only to those entitled to them.166 In 2009, the MoI was still labouring to validate the eligibility of about 37 per cent of its force, mainly due to the lack of cooperation of three police zone commanders.167 Perhaps one of the greatest constraints in enforcing discipline within the police was the constant failure to prosecute deserters. Under Minister Mohammadi, a new Act of Duty was introduced in 2010, which asked each new police officer to commit himself to ten years of service, while NCOs (non‐commissioned officers) would have to serve five and patrolmen three. They could, from that point onwards, be prosecuted if they left earlier. The punishment was however not specified and President Karzai appeared to be on a different track as he kept proclaiming amnesties for deserters.168 As of 2011, discipline remained a major problem within the MoI. After his appointment in 2010, Minister Mohammadi had to start travelling to the provinces himself to understand what was going on and convey that the behaviour of the police was under scrutiny. What he found often surprised him negatively.169 5.5 Control: Internal affairs The role of internal affairs is crucial in stemming the tide of corruption and abusive behaviour of any police force. Initially the capacity of the MoI internal affairs department (called Office of the Inspector General) was very modest. At each year’s end Kabul used to despatch officers to the provinces to assess their work. The departments of counter‐terrorism, counter‐narcotics and a few 165
Rosen, Something (see FN 160). ‘Afghanistan Security’, 25 (see FN 93). 167
Ibid. 168
Meeting with NTM‐A officers, November 2010; ‘More Afghans Deserting Army, but Karzai Turns ‘Blind Eye’, The Washington Post, 3 September 2011. 169
Personal communication with NTM‐A personnel, November 2010. 166
Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context more had their own inspecting department.170 In practice, very little was going on in terms of seriously investigating police abuses. The Office of the Inspector General was first strengthened in 2008, when sections were established in each of the six regional commands. These sections were in charge of inspecting the provinces and were typically staffed by one head and 11–12 officers. Although this meant that a steadier form of supervision from the centre was now in place, at least in principle, its impact was still limited, not least because the MoI still showed little willingness to prosecute misbehaving police: a police officer serving in Wardak reported cases of corruption and abuse by the police to internal affairs, but the delegation which collected the information did not take any action.171 In fact many investigations based on internal affairs’ efforts mostly targeted low rank policemen rather than senior officials (see also Section 4 above).172 Further developments meant to strengthen the role of internal affairs included the establishment in 2010 of a commission to inspect the provinces, focusing on logistics. The anti‐corruption unit within the office was developed into six teams connected with Interpol, tasked to travel to the provinces. Composed of six members each and equipped with special equipment by the British, the teams started work in mid‐2010. The major crime task force was also supposed to assist the Inspector General.173 Some of the MoI reformists dismissed such teams as more of a propaganda exercise than anything else.174 What is certain is that in 2010, internal affairs was still not the focus of much effort: it still did not have any written procedures for inspections and audits and its staff never received formal training. They were only getting on‐the‐job training from older inspectors.175 Somewhat surprisingly, the Office of the Inspector General started only being mentored by NTM‐A in the second half of 2010.176 Aside from the inspecting teams sent from Kabul, two regional (zone) organisations also perform some type of oversight over the provincial police, in particular prosecuting those police reported to 170
Meeting with NTM‐A officer, November 2010; interview with Haidar Bassir, former deputy minister, Kabul, 31 October 2010. 171
Interview with MoI official, Kabul, April 2010. 172
International Crisis Group, Reforming, 13 (see FN 57). 173
Meeting with NTM‐A officer, November 2010. 174
Interview with senior MoI official, November 2010. 175
NTM‐A, ‘90 Day Assessment’, 1 October 2010. 176
Interview with foreign diplomat, November 2010. them by the inspecting teams or any other source. They are the Military Attorney Directorate (Tsaranwali Nezami) and the Military Court (Mahkameh Nezami).177 The Military Attorney Directorate is responsible for investigating and prosecuting the cases of crimes perpetrated by the military personnel – the police as well as the army. The cases processed by the directorate constitute three types: One, cases are referred to the Military Attorney Directorate by the police or the army when they discover crimes committed by their personnel. Second, a case may be initiated as ordinary citizens file complaints against the police or army personnel. Third, the Detective Unit of the Directorate may visit the police or the army and discover crimes perpetrated by them.178 In any of these cases, the directorate is required to complete the investigation of the case within 15 days and send it to the Military Court.179 Section 6 illustrates how these two organisations function in practice in the case of one specific province, Herat. Both were obviously understaffed and unable to cope with the workload and faced political pressure when investigating cases. In sum, the development of internal affairs’ capacity to deal with police misbehaviour started in 2008, but was never given much priority. Dedicated units were formed, but never staffed adequately, both in qualitative and quantitative terms. Their ability to thoroughly investigate allegations of misconduct was usually insufficient to lead to successful prosecution, not to speak of the problem of political interference. 5.6 Conclusion Efforts to improve command and control after 2002 narrowly focused on getting the police in better shape for the counter‐insurgency effort. This had major implications for the functionality of the system. Significantly, internal affairs, more relevant to civilian policing, was neglected throughout the 2002–10 period and had only minimal impact on the functioning of the police. The technical capability to manage the fighting was greatly increased, but the willingness to fight had to be ensured through the appointment of individuals committed for personal or political reasons. The same individuals, however, most of 177
Interview with officials from Herat Military Court, September 2009. 178
Interview with officials from Herat Military Attorney Directorate, September 2009. 179
Interview with officials from Herat Military Attorney Directorate, September 2009. November 2011 27
28 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
the time showed little or no commitment to maintaining good relations with local communities, containing corruption and operating a transparent system. Much of the difficulty in setting up a functional chain of command and control at the MoI might have been self‐inflicted pain, not only because it developed the tactical ability to fight while neglecting all other aspects of policing. The model adopted, styled after the European bureaucratised‐centralised version of a police force, is very demanding in terms of human resources. It requires several layers of supervision and monitoring to work, all of which have to be staffed by well‐trained and relatively highly educated officers. The system has to produce a paperwork trail for supervisors to follow, which means that even in the field, senior and middle rank police officers need to be educated to the level where they can produce meaningful reporting and recording. It is doubtful whether the MoI will be able to mobilise such skilled human capital in the foreseeable future, given that it is in short supply in Afghanistan in general and that the MoI has consistently failed to mobilise human resources available on the labour market. The MoI hardly recruits in Afghanistan’s cities. If we were to assume that a balanced version of the Western‐Turkish model adopted by the MoI was not feasible in the post‐2001 context, perhaps looking for alternative models would have been better than developing a wholly unbalanced one. The so‐called Anglo‐Saxon model of policing relies more on local accountability, making several levels of bureaucratic supervision less necessary. Perhaps in the Afghan context some form of local policing, embedded in the communities and responsible to them, could have worked. An adviser noted that although the police usually had no written orders or records and relied on an oral system, they had a formidable memory and could handle their tasks. This could be considered a basic form of professionalism in its own way.180 However, while in some cases the police might strike a good relationship with local communities (being drawn from them) or with local authorities (based on shared professionalism or, more likely, a shared political or ethnic background), seeing this pattern repeated throughout the country seemed unlikely in 2011 Afghanistan. As discussed in Section 4.2, the main problem is the fragmentation of the country into often‐rival communities, even within a single district; strong police connections to 180
Interview with foreign adviser, Kabul, October 2009. AAN Thematic Report 07/2011 a particular community would often come at the detriment of relations with other communities. Some experiments in attributing an oversight or advisory role to local councils were going on in 2010–11 in some parts of Afghanistan, as part of a Democrating Policing project sponsored by UNDP, the Canadians, the British and others, but at the time of writing it was too early to say whether these experiments had the potential to offer an alternative model.181 The 2005 and 2010 parliaments were both quite keen on exercising oversight over the MoI,182 but had little support from other institutions of the Afghan state and limited legitimacy of their own, particularly in the case of the 2010 parliament, given the extensive rigging which took place during its election. 6. HERAT: A CASE STUDY OF A PATRIMONIAL COMMAND AND CONTROL This paper concludes with an in‐depth study of Herat’s police, illustrating how the exercise of command and control has played out in a particular province. This paper draws examples and anecdotes from a number of Afghan provinces to document developments within Afghanistan’s police, a methodology which has both positive and negative implications. The main risk implicit in it is to handpick information to suit a particular point of view or interpretation of events. It seemed therefore appropriate to apply the framework adopted in this paper to a particular province and discuss that province in detail, testing the validity of the framework. Although Herat is not necessarily representative of other Afghan provinces, its case illustrates the shape command and control can take, away from the capital. 6.1 Police during Ismail Khan’s governorship (2001–4) With the collapse of the Taleban in late 2001, Ismail Khan, who had ruled Herat between 1992 and 1995, once again was able to consolidate his control over Herat. The political dynamics of Afghanistan were different than during his first rule, when the centre was barely able to interfere. The Karzai government, supported by the US and the international community, gradually started 181
Personal communication with Heather Coyne of NTM‐A, Kabul, November 2010. 182
Personal communication with foreign diplomats in Kabul, April 2011. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context taking on the warlords who ruled the provinces, particularly those like Ismail Khan who did not have strong supporters in Kabul.183 As a result, Ismail Khan was more responsive to popular demands and tried to attain popular support in Herat so that the central government would find it more difficult to unseat him.184 To this end he tried to keep his subordinates under control so that they would have less space to prey on the population: he held public session regularly in his governor’s office to receive petitions, and used his personal authority and power to follow them up; he also adopted a micro‐management style and strict personal oversight of the government staff to keep them accountable. This was in contrast to his style of government in the 1990s, when he left greater room for manoeuvre to his subordinates. Like his peers in other regions of Afghanistan, he derived his system of command and control from his experience as a guerrilla commander, as well as – more superficially – from his background as military officer. As his former subordinates put it, rather than relying on formal institutions, a commander of guerrilla fighters must rely on personal oversight and supervision to control the fighters.185 This attitude reflected the unwillingness of Ismail Khan to rely on party organisation and party structures in the 1980s, even when he formally belonged to a party (Jamiat‐i Islami) which had political cadres in Herat.186 Although his strategy undermined the development of formal rules and procedures, and therefore institutions, in the short term it was effective in keeping public employees under control and limiting their abuses. This was true also of the police.187 Being heavily dependent on Ismail Khan’s charisma and financial wealth, the system was extremely vulnerable to any change in the fortunes of the leader. It was also bound to attract the attention of centralists in Kabul, to 183
Gulshan Dietl, ‘War, Peace and the Warlords: The Case of Ismail Khan of Herat in Afghanistan’, Turkish Journal of International Relations 3, Nos. 2&3 (2004): 41–66; Antonio Giustozzi, Empires of Mud (London: Hurst, 2009). 184
Interview with a MP from Herat, Herat, August 2009. 185
Interviews with former militia commanders; interviews with provincial police officers, Herat, June‐
September 2009. 186
On this point see Antonio Giustozzi, ‘The Missing Ingredient: Non‐ideological Insurgency and State Collapse in Western Afghanistan, 1979–1992’, Working Paper Series 2 n. 11 (London: Crisis States Research Centre, LSE, 2007). 187
Interviews with MPs from Herat; interviews with provincial police officers, Herat, June‐September 2009. whom Ismail Khan’s personal power represented a potential strategic threat. 6.1.1 Human resources Unlike his first period of rule over Herat (1992–5), Ismail Khan did not implement conscription after 2001. To staff the police force, he relied on his militia commanders and their networks. He or his son, for instance, would recruit former militia commanders who had fought with them against the Russians or the Taleban and appoint them to key positions within the provincial police headquarters or as heads of police stations. These commanders then would recruit their former fighters to work as their subordinates within those police units. The relationships between the commanders and Ismail Khan as well as between the commanders and their fighters were based on personal trust and loyalty, although in practice guaranteeing the reliability of the individuals chosen was not always possible, as political alliances were still necessary.188 Ruthless enforcement of Ismail Khan’s own rules was his recipe for keeping order in the police force. The fate of Khawaja Issa, whom Ismail Khan had appointed as the district police chief in Fourth District, illustrates Ismail Khan’s method. Khawaja Issa belonged to a faction that had problems with Ismail Khan in the past (the Afzali Front). District Four included the commercial zone of Herat city and the exchange market. In early 2003, money traders complained to Ismail Khan about instances of night robberies from the exchange market. Ismail Khan put pressure on Khawaja Issa to make sure that the commercial centre was secured, but the situation did not improve, and the traders accused the district police of being linked to the thefts. He then personally sent an order, dismissing Khawaja Issa from his post. Khawaja Issa, however, resisted the order, claiming that it had to be issued by the provincial police chief, Mama Zia, rather than Ismail Khan. Since Khawaja Issa had family ties to Mama Zia, he hoped that he could retain his position with Zia’s support. Facing Khawaja Issa’s defiance, Ismail Khan then went to the district police station with his own body guards, and his effort to disarm Khawaja Issa led to a violent clash between their men. Khawaja Issa became a fugitive and left Herat. This sent a warning to other police officers that their survival depended on 188
Interviews with provincial police officers; interviews with a political analyst, Herat, June‐September 2009. November 2011 29
30 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
following Ismail Khan’s command.189 As a police officer commented, The police under Ismail Khan was more accountable since they were his own men. He knew them personally and exerted tight control over them. Police officers knew that they owed their position to obeying Ismail Khan.190 personal decisions. The district police chiefs would usually report to Ismail Khan and after detaining someone, the police chiefs often consulted Ismail Khan personally and followed his order.193 Similarly the officials from the military court complained that under Ismail Khan, formal procedures were not observed. An official from Herat Military Court stated that The military and civilian courts were not separated; the official rules and procedures were not observed. Civilian cases were often sent to us since the civilian courts were overwhelmed. The offences of the police and army personnel, on the other hand, were often punished by their superior or Ismail Khan, without referring such cases to the Military Attorney Directorate and the Military Court.194 6.1.2 Supervision and oversight Even more important and effective than repressing open disobedience were Ismail Khan’s random personal visits to the police. He often patrolled the city in person after midnight – sometimes on horseback – to oversee the police, even his confidants, and to punish officers who did not take their jobs seriously. Many police officers who served during his rule stated that the police were afraid to ask for bribes openly or extort from the population out of fear of punishment by Ismail Khan. Although Ismail Khan’s personal oversight of the police force made it relatively accountable, his approach undermined the formal rules and procedures as well as the development of institutions, hence long‐term stability. His use of the police to his own advantage and to that of his political faction also stirred opposition among groups that soon started seeking support in Kabul.191 Formal rules of reporting were gradually being established by the MoI, and the provincial police headquarter was asserting its role at the top of the provincial police, but the procedures were often bypassed by Ismail Khan’s orders and the reporting system and due process were not established.192 Strict oversight limited the ability of the police to collect bribes and extort from the population, but did not stop corruption. The underdevelopment of systematic reporting and surveillance was a major limitation. Since Ismail Khan had to rely on a small number of trusted loyalists, the extent to which his confidants could monitor and oversee the police was limited. As a result, the senior police officers could exploit the weakness of the reporting system to skim off public resources. As a police officer highlighted, under Ismail Khan it was much harder for the police to ask bribes from the population, but the senior police officers did have a lot of payda [informal and mainly illegal revenue]. The few senior officers at a district police station, for instance, could collude with each other and inflate the number of personnel that they had. They kept two lists: one with real number of the personnel and the other list including the ghost personnel. When sending the list to the provincial centre for salaries and food, they would send the inflated list and then distribute among themselves the extra money and the food they received from the centre. When Ismail Khan’s men visited the police stations to check on the attendance, they would use the accurate list without ghost personnel.195 District police stations are required by the internal rules of the police to report regularly to provincial police headquarter. They must file a report if they detain an individual for breaking the law or committing a crime, then send the suspect and the report to the CID of the provincial police headquarters. The CID then must complete the file and send it to the provincial attorney directorate. Under Ismail Khan, however, these procedures were often subverted. Subverting rules was common throughout Afghanistan at that time, but in Herat’s case, what was peculiar was the police force’s heavy reliance the police on Ismail Khan’s 189
The story of Khawaja Issa was narrated by many provincial police officers, Herat, July‐September 2009. 190
Interview with a senior police officer from regional police headquarters, Herat, August 2009. 191
See Giustozzi, Empires (FN 183). 192
Interviews with police officers who served under Ismail Khan, Herat, August 2009. AAN Thematic Report 07/2011 193
Ibid. Interview with official from Herat Military Court, Herat, September 2009. 195
Interview with a deputy district police chief, Herat, August 2009. 194
Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context The underdevelopment of the reporting and analysis system meant that such misrepresentations often went unnoticed and made the system vulnerable to exploitation. Ismail Khan’s system however guaranteed territorial and population control, as well as basic security for common citizens. The staffing of the provincial police did not have to follow any formal limits or restrictions. The number of police personnel in police stations and substations were flexible – they could be increased or decreased as deemed appropriate – given Ismail Khan’s discretion over paying police salaries from the provincial custom revenues.196 Ismail Khan’s police system could be described as effective in delivering what the MoI of 2011 seemed to consider its priorities: paramilitary control of territory and population (see Section 5. It was also effective in providing basic security to the population because it was managed at the provincial level by Ismail Khan himself, someone who had a strong sense of ownership over the system. In other words, Ismail Khan had a strong interest in overseeing the police. The main problem with this form of patrimonially‐led policing is that few of Afghanistan’s strongmen had the self‐confidence and the staying power of Ismail Khan and therefore were not motivated to invest as much in making the system functional. Moreover, Ismail Khan never managed to secure the fringes of Herat province (such as Shindand district), where his personal rule aroused hostility and the police force was unable to penetrate.197 A degree of conflict was therefore built into the system. In addition, in Ismail Khan’s case the patrimonial character of the system was incompatible with the rule of law (that is with a set of rules applied independently of the whims of individuals), even if an evolution in such a direction could not be ruled out. Ismail Khan himself was the law and could arbitrarily determine any case. This was always going to make such an option unpalatable to Europeans at least and probably to the US Department of State too. system of personal rule and altered how the police operated in Herat. The central government gradually tried to restructure the police force in Herat, in line with the reforms going on nationwide, with the goal of ending its loyalty to Ismail Khan and including it in the Afghan National Police.198 First, all police officers had to pass through a police reform program to remain in their jobs. Second, the central government took control of police appointments, gradually appointing officers from outside Herat to senior positions within the provincial police force.199 As in other provinces, therefore, the Herat police force received an injection of non‐local police, to mix with a local force composed of both supporters and – increasingly – opponents of Ismail Khan. This can be described as a surrogate form of centralisation (see Introduction); the evidence suggests that it did not improve command and control. The central government gradually removed Ismail Khan’s loyalists from senior police posts or transferred them to other provinces. Instead police officers or sometimes militia commanders loyal to other factions were appointed to senior police posts in Herat. The provincial police force, therefore, over time fragmented internally among different factions.200 The diversity of political affiliation of the police had two major consequences: First, the police had greater impunity than under Ismail Khan. Many of the new officers belonged to pro‐central government political factions and the central government intended to preserve their support. As a result, these police enjoyed more freedom and greater impunity.201 Factionalism reinforced conflict and rivalry within the police force and undermined its cohesion. Since the survival of police officers depended on the strength of the political factions they belonged to, officers tried to strengthen the position of their faction within the police and MoI while undermining other factions. This factional competition exacerbated mistrust and undermined cohesion and cooperation. For instance, in many 6.2 Police in post–Ismail Khan Herat 6.2.1 Synchronisation with Kabul The removal of Ismail Khan from his post as the governor of Herat just one month before the first presidential election (October 2004) ended his 196
Interview with senior police officer from regional police headquarters, Herat, September 2009. 197
On this see Giustozzi, Empires (FN 183). 198
To be included in the Afghan National Police and receive their full salary entitlements, local police forces had to undergo training imparted by US Department of State contractors and register the policemen with the MoI. 199
Interviews with officials from Herat Police Training Center, Herat, October 2010. 200
Interviews with provincial police officers, Herat, August‐September 2009. 201
Interview with a senior provincial police officer, August 2009; interview with a provincial government official, August 2009. November 2011 31
32 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
cases a district police chief and his deputy belonged to different factions and, instead of working as a team, they refrained from supporting and cooperating with each other.202 In conclusion, the removal of Ismail Khan from Herat’s governorship in 2004 did not unleash a new phase of institution building, but instead saw the introduction of fragmented patrimonialism, dominated by short‐term concerns of various small actors, none of whom had enough control over the environment to develop the capacity or even the interest in improving the performance of the police. Moreover, police officers opposed to Ismail Khan still had little interest in implementing orders coming from Kabul, the more so since the MoI was hardly able to monitor their operations. 6.2.2 Supervision and oversight from Kabul With the inclusion of the Herat provincial police force in the Afghan National Police, command and control also changed. The Ministry of Interior had to rely on reports sent by officials stationed in Herat. Each of the MoI directorates sent inspecting teams every few months to oversee the provincial police and report back. The central government, however, faced major challenges in controlling the provincial police through these two mechanisms. The inspecting teams, supposed to function as the agents of the MoI, enjoyed discretion and monopoly over the transfer of information from Herat to Kabul. This allowed them to abuse their authority and manipulate information they relayed. The spatial and temporal distance from Herat to Kabul undermined the MoI’s capability to exert the same control over the police that Ismail Khan had enjoyed. The provincial police officials often treated the inspecting teams very well. They were taken to luxurious hotels, offered good receptions during their visit to Herat, and even offered expensive gifts before their return to Kabul. In return for such a welcome, they would often send a very positive evaluation of the provincial police to Kabul.203 This process exemplifies the standard principal‐
agent dilemma in which the principal is not able to monitor the agent to ensure the accuracy of the information it receives. Those in control of the police force after Ismail Khan often had a short horizon of remaining in power. Their goal was to maximise the resources they could extract from the population, and to channel part of those resources 202
Interviews with provincial police officers, Herat, July‐
September 2009. 203
Interview with a senior provincial police officer, August 2009; interview with an official from the provincial police guest house, Herat, September 2009. AAN Thematic Report 07/2011 back to the political patron on whom they depended for their current position as well as to future appointments. In other words, higher‐ranking officers sold lower‐ranking posts. Police officers started to demand advances from lower‐ranking officers and soldiers, as well as regular payments, to remain in their jobs. An example frequently cited by provincial police officers, to show the hierarchical nature of corruption, was that the provincial police chief dismissed the head of Ghor Darwaz police sub‐
station in northern Herat for failing to make the monthly payment of Afn 50,000 (at the time roughly US$ 1,000).204 The lower‐ranking officers and soldiers collected bribes from the population to make payments to senior officers. Not surprisingly, in post‐Ismail Khan Herat, petty corruption within the police force and predatory activities by the police increased drastically. Almost all ordinary citizens as well as police officers who were interviewed in Herat believed that police corruption – particularly collecting bribes from the population – had soared in post‐Ismail Khan Herat.205 This pattern appears to have been repeated in other Afghan provinces as well. As pointed out in Section 5.5, Internal Affairs continued to play a small role in containing corruption within the police. External responsibility for pursuing corruption and misbehaviour within the police falls to the Military Attorney Directorate. In Herat, the directorate was short of staff as of 2010 (as it probably was elsewhere, too). With a staff of 15, it was responsible for all the cases in the Western Zone, which includes four provinces – Herat, Farah, Ghor and Badghis. According to senior officials from the directorate, they did not have enough manpower to process the files referred to them by the police, let alone investigate them.206 Such cases often involved abuse of public property, such as police cars or weaponry. In spite of the increased involvement of the police in demanding bribes from the population, cases of corruption were rarely referred to the directorate.207 The latter rarely had the time to form detective units to visit the police and oversee their performance. According to a senior official from the military court, they did not prosecute even a single case of police corruption from the four provinces over the September 2008– 204
Interviews with provincial police officers, Herat, September 2009. 205
Interviews with police officers, government officials, and ordinary citizens, Herat, July‐September 2009. 206
Interview with senior officials from Military Attorney Directorate, Herat, September 2009. 207
Interview with a senior official from Military Attorney Directorate, Herat, September 2009. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context Table 2. Illegal tolls collected by Herat police from truckers, and level of oversight
Area (distance from Herat city in kilometres) Frequency of illegal toll collection per trip Amount collected at each toll (AFNs) Level of supervision and oversight Iranian border to Herat city (120 km) 2 50
Medium
Herat City 1 20
High
Herat Provincial Police Headquarters or regional police HQ Oversight organisation Border police Herat City to Farahrud (Farah Province) (435 km) 12 20
High
Shindand Police Headquarters or regional police HQ Farahrud to Delaram (Nimruz Province) (545 km) 17 80–150
Low
Delaram Police Headquarters Delaram to Gereshk (Helmand Province) (655 km) 0 0
High
Afghan National Army Source: Interviews with truck drivers and merchants, Herat, November 2010 208
September 2009 period. During the September 2007–September 2009 period, the military court did not receive any case of police corruption from Badghis province. According to the same source, when the Mujahedin controlled the provincial government [referring to Ismail Khan’s faction], the police chiefs sent us many cases where their subordinates were accused of corruption and demanding bribes. Nowadays, however, the police chiefs rarely send us cases of corruption. It is evident that corruption exists, but I believe either the police chiefs prefer not to report such cases or there is not effective oversight of the police.209 Furthermore, the directorate officials claimed that their efforts were sometimes undermined by the threats that they received or pressure from ‘the top’ when prosecuting crimes committed by the police or the army. According to a senior official of the directorate, under pressure from influential political figures, they sometimes had to halt investigations and send incomplete cases to the military court. The military court, then, would dismiss such cases due to lack of evidence.210 In addition, the directorate’s oversight of the police was further undermined by the threats that its personnel received from supporters of those involved in the case.211 6.2.3 The impact of supervision on illegal tolls The lack of oversight resulted in corruption and illegal behaviour expanded rapidly, often in obvious forms. To gain insights on the extent to which the police engage in extortion and what factors could impact police accountability, a study of illegal tolls collected by the police along the ring road was conducted in November 2010. The ring road starts at the Iranian border in the west of Herat; passing through the southern, eastern and northern provinces, it connects the entire country. The interviews were conducted with twelve truck drivers who transported goods along the ring road. The accuracy of the information provided by truckers was cross checked in another five interviews with traders who sent goods from Herat to Kabul or Mazar‐e Sharif in the north. 208
Interview with an official from Herat Military Court, Herat, September 2009. 209
Ibid. 210
211
Ibid. Ibid. November 2011 33
34 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
A review of illegal tolls collected along the ring road passing through Herat shows that the extent to which police engage in extortion is inversely correlated with the scope of oversight of the police: the less effective oversight is, the more extortion the police engage in. The further away from the provincial centre the police are, the higher the amount of illegal tolls they collect from truckers and the more abusive they are. The major government and police institutions – including the provincial police headquarters and the regional police headquarters – are in Herat city; as the distance from Herat city increases, these two institutions are less likely to patrol the roads and oversee the police substations along the road. This allows substations in remote areas to extract more from truckers. As indicated in Table 2, the amount of tolls collected by the police is smallest in Herat city and along the road from Herat to Farahrud – usually around 20 Afghani from each truck.212 The random visits by inspecting teams from the regional police headquarters in Herat city limits the freedom of these police substations to engage in extortion from truckers. From Farahrud to Delaram, truckers pay the highest tolls and are often subject to the most predatory activities and abuses by the police. The police along this route are monitored only by their own supervisors in the Delaram Police Headquarters and not by the regional police headquarters. Truckers suspect collusion among the police substations and their supervisors since frequent complaints by truckers to the Delaram Police Headquarters has not reduced the predatory activities of the police. The only time that the police in this region stop extorting from truckers is when the Afghan National Army patrols the road.213 An anecdote by a trucker depicts the importance of this external oversight: When the National Army cars appear on the road, the police stop harassing truckers and demanding money. In many instances, I was stopped by the National Army patrols and asked whether the prior police substation had forced me to give them money. I told them that I was abused and forced to pay. Later I heard from other truckers that that police personnel in that substation were punished by the National Army and were disarmed. The police do fear the National Army and do not ask for money when they hear that the National Army patrols the road.214 212 Interviews with truck drivers, Herat, November 2010. Ibid. 214
Ibid. 213
AAN Thematic Report 07/2011 From Delaram to Gereshk, the road is the under control of the National Army. While along other parts of the ring road police cars may establish temporary checkpoints and collect illegal tolls from truckers in addition to what is collected by stationary police substations, between Delaram and Gereshk, such temporary police stops are rare. The control of the National Army over this interval means that truckers do not pay any tolls. The National Army neither collects tolls from truckers nor allows the police to establish temporary stops and collect money from truckers.215 From the discussion of oversight in the last two sections it emerges that in the rural areas very few institutions or organisations exist that are capable of overseeing or supervising the police outside the MoI: probably just the Afghan army, provided it is positioned to do so. For this reason, the MoI’s own command and control (of which oversight and supervision are aspects) are of high importance. Without that, any gain in tactical proficiency, firepower, mobility and communication capabilities is not likely to produce much benefit even for the Afghan state. 7. CONCLUSION Afghanistan’s police force has changed significantly from what it was in 2002. When looking specifically at command and control, the technical capacities of the MoI have clearly improved dramatically since 2002. Direct communication between the MoI and the zones and the provinces is much easier now. Reporting from the bottom to the top, almost non‐existent in 2002, existed in some measure in 2011: the technology to report incidents was there and mostly functioning. The technical ability for command and control, however, does not imply that there is a willingness to command at the top, or the ability to direct effectively, or that the bottom is willing to obey. Various forms of indiscipline and even collusion with the insurgents continued to be common in 2010, even if some efforts (surprise visits by the minister to stations and check posts) were going on in the second half of the year to reduce the incidence of such behaviour. Undoubtedly, the capability of the MoI to supervise its structures on the ground has been slowly improving in 2010, again from virtually zero. More and more policemen have been the object of disciplinary and corruption enquiries, even if prosecution remains very rare; most importantly, political connections 215
Ibid. Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context still allow high rank officers to get away with criminal behaviour and, in general, the ability to process the thousands of enquiries is very limited. weak institutionalisation of the MoI means that that the question of what will happen in the event of Mohammadi’s departure always looms. The Ministry of Interior of 2011, therefore, is still half way between the Rabbani government of the 1990s and the Daud Khan government of the 1970s. The effort is to re‐establish some kind of ‘moderate meritocracy’ within the system, as it had existed under Daud and the leftist governments of the 1980s, but the effort has to coexist with much heavier doses of political patronage and a degree of corruption, which is where it resembles the system of the 1990s. Every change of minister has brought changes in personnel, with major disruptions in the functioning of the patrimonial machinery of the MoI, bringing uncertainty among key staff and removing incentives for long‐term planning of any type. Only corruption was not disrupted; quite the contrary, it was encouraged as financial rewards had to be accumulated on a shorter timescale, before a new wave of appointments. Mohammadi may not stay in his job, meaning that the recent (modest) gains of the MoI as a fighting machine rest on shaky ground. In the 1970s, the political regime was more solid and did not face significant challenges until 1978, while the opportunities for corruption were very modest because of the small size of the illegal economy and high levels of state control over wealth. In the 1980s, the open challenge to the government was stronger than ever before in the history of the Afghan state, but at least the regime was based on a political party of some strength, which allowed for oversight of the police through a capillary structure of party organisations; this played an important role in maintaining discipline and containing corruption, even if the charismatic role of Gulabzoi as minister of interior should not be underestimated. In the 1990s, all that was in place was a relatively loose network of local military leaders, personally loyal to the political leadership of the time (Rabbani and Massud); the ideological dimension of affiliation to their party had by then evaporated and the organisational structures of Jamiat‐i Islami had always been extremely weak. In 2011, the MoI could not rely on any political organisation, despite being heavily politicised. Political factions organised themselves in networks, as they had in the 1990s, without any formal structure and often‐shifting composition and alliances, making them unsuitable for regular oversight. The very high level of corruption turned into a major obstacle tore‐establishing a moderate meritocracy within the system. The current practice of relying on former comrades in arms of Minister Mohammadi to wage the counter‐insurgency had some military impact, but also amounts to a revitalisation of a patrimonial system rather than the strengthening of the institutional machinery of the MoI. This re‐
patrimonialisation is divisive, as it alienates those who are not favoured by the minister. Moreover, like every patrimonial system it depends on a few charismatic personalities, who might die (as quite a few of Mohammadi’s appointees have) or lose the political favour of the ruling elite. Indeed, the very Although the MoI has moved towards strengthening the paramilitary character of the police after 2002, it has not done so in a coherent, planned way. It has just drifted in that direction, driven by external developments (the insurgency). The result is a mismatch between the implicit aim of a paramilitary force (which is how the bulk of the police is being trained) and the capacity of the MoI to put that into practice by turning the police into an effective, well‐coordinated force. There is mileage in the Afghan police system to be gained by strengthening supervision, oversight and monitoring and the gains could be relatively quick. It is not clear, however, who is going to do it in a police force where patronage and nepotism still predominate over professionalism, political organisation does not develop beyond the form of informal networks and top leadership is precarious and subject to frequent turnover. No minister of interior has been in power for more than three years after 2001; this is too little for any patrimonial system to reach full steam. In some areas, Afghanistan has had a de facto de‐
centralised (in fact almost feudal) model of policing, as illustrated in the chapter on Ismail Khan’s Herat (current examples can also be found in Balkh or Kandahar). The model seems to represent a better match for the human resources that exist within the MoI, as it can function with a limited number of professional or educated officers. As shown in the case of Herat, however, this might do as much damage as bring benefits and still requires a capable centre to manage it effectively. Furthermore, this is not likely to be the solution to an insurgency which has grown beyond the stage of localised revolt. Finally, charismatic leadership is always in short supply and cannot be ‘manufactured’: training and education can form professional officers, but not charismatic leaders. November 2011 35
36 Giustozzi and Isaqzadeh: Afghanistan’s Paramilitary Policing in Context
The model, even if it works to some extent, cannot be easily replicated. Alternative approaches to managing an organisation like the MoI include institution building, of which there is still very little sign; reliance on ideology (nationalist or otherwise), of which there is not even talk; and external oversight, which in 2011 was still at the experimental stage. In the absence of any strategy to lead the MoI towards a kind of ‘moderate meritocracy’, the MoI as an institution will remain brittle and subject to sudden, dramatic crises. It is not going to play the role of paramilitary organisations like the Turkish gendarme, the Italian Carabinieri or the French Gendarmerie in state building. The local crises that the police force currently faces, for instance in terms of indiscipline, insubordination or collaboration with the enemy, all occur in a military and political AAN Thematic Report 07/2011 environment ‘armoured’ by the strong presence of allied foreign military forces. One cannot help wondering what would happen in their absence. In this regard, the mere numerical expansion of the police is not likely to represent much of a substantial strengthening and is even likely to seriously complicate the task of developing an effective command‐and‐control structure, as already limited skills available will be further diluted. While signs are few of a system of promotion and appointment that could quickly produce a new generation of capable officers, awareness of the necessity for progress in this direction is growing within the international community and a sense of urgency is emerging. Within the Afghan MoI, however, the predominant concern remains political infighting and meritocracy is still mostly paid lip service only. ABOUT THE AFGHANISTAN ANALYSTS NETWORK (AAN) The Afghanistan Analysts Network (AAN) is a non‐profit, independent policy research organisation. It aims to bring together the knowledge and experience of a large number of experts to inform policy and increase the understanding of Afghan realities. The institutional structure of AAN includes a core team of analysts and a network of regular contributors with expertise in the fields of Afghan politics, governance, rule of law and security. AAN will publish regular in‐depth thematic reports, policy briefings and comments. The main channel for dissemination of the reports is the AAN web site. For further information, please visit www.aan‐afghanistan.org. ABOUT THE AUTHORS: ANTONIO GIUSTOZZI AND MOHAMMAD ISAQZADEH Dr. Antonio Giustozzi is an independent researcher associated with IDEAS (LSE). He is the author of several articles and papers on Afghanistan, as well as of three books, War, Politics and Society in Afghanistan, 1978‐
1992 (Georgetown University Press), Koran, Kalashnikov and Laptop: The Neo‐Taliban Insurgency, 2002‐7 (Columbia University Press) and Empires of Mud: War and Warlords in Afghanistan (Columbia University Press), as well as a volume on the role of coercion and violence in state‐building, The Art of Coercion (Columbia University Press, 2011). He also edited a volume on the Taliban, Decoding the New Taliban (Columbia University Press, 2009), featuring contributions by specialists from different backgrounds. He is currently researching issues of governance in Afghanistan from a wide‐ranging perspective which includes understanding the role of army, police, sub‐national governance and intelligence system. Mohammad Isaqzadeh is Assistant Professor of Political Science at the American University of Afghanistan and works as a consultant for the World Bank in Kabul. His primary research interests include the relationship between religion and electoral politics, politics of corruption and policing in Afghanistan. He received his MPhil in International Development Studies from the University of Oxford in 2010 and his BA in political science from UC Berkeley in 2008. © Afghanistan Analysts Network 2011. All rights reserved. No part of this publication can may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without full attribution.